Radionuclide scintigraphy - gamma-ray imaging of a radioindicator in nuclear medicine

AstroNuclPhysics ® Nuclear Physics - Astrophysics - Cosmology - Philosophy Physics and nuclear medicine

4. Radionuclide scintigraphy
4.1. The essence and methods of scintigraphy.
4.2. Scintillation cameras
4.3. Tomographic scintigraphy
4.4. Gated scintigraphy
4.5. Physical parameters of scintigraphy - image quality and phantom measurements
4.6. Relationship between scintigraphy and other imaging methods
4.7. Mathematical analysis and computer evaluation in nuclear medicine
4.8. Radionuclides and radiopharmaceuticals for scintigraphy
4.9. Clinical scintigraphic diagnostics in nuclear medicine


4.1. The essence and methods of scintigraphy

Radionuclides in nuclear medicine
Nuclear medicine is a field dealing with diagnostics and therapy using open radioactive substances - radiopharmaceuticals - applied to the internal environment of the organism; these in vivo methods will be addressed in this chapter. In an in vitro test, the radiopharmaceutical is not administered to the patient's body, but is used in the radiochemical analysis of blood samples taken; the patient does not come into contact with a radioactive substance, only a sample of plasma or other body fluid is used (in vitro radioisotope methods are briefly outlined in §3.5 "Radioisotope tracking methods", passage "Radioimmunoassay - radiosaturation analysis"; now they are mostly not part of nuclear medicine, but laboratory biochemistry).
  Nuclear medicine methods are based on two basic properties of radionuclides : 
1. Emissions of penetrating ionizing radiation during radioactive transformations of nuclei (detailed physical explanation in §1.2"Radioactivity") ;
2.
I
dentical chemical behavior of isotopes Þ radioactive isotopes react chemically in exactly the same way as stable isotopes of the same element (§3.5 "Radioisotope tracer method") .
   Radioactive atoms and their molecules - compounds "labeled" with radioactive elements - are distributed in the body as if they were non-radioactive, but penetrating radiation is continuously emitted during the radioactive transformation of the respective nuclei. This radiation allows them to be "made visible" - to monitor, indicate, "trace" *) - and measure their amount by detection devices during diagnosis, or the radiobiological effects this radiation can be used for therapeutic purposes.
*) Hence the general name of an indicators or tracers methods, that are used not only with the help of radionuclides and not only in medicine, but also in laboratory and industrial applications (§3.5 "Radioisotope tracking methods") .


Scintigraphic diagnostics and radionuclide therapy in nuclear medicine

The central method of nuclear medicine is radioisotope diagnostics in vivo: we apply a suitable (bio)chemical substance with a bound radionuclide - the so-called radioindicator or radiopharmaceutical - to the organism. This substance enters the metabolism and is distributed in the body according to its chemical composition - pharmacokinetics - of the given radioindicator. Physiologically or pathologically, it accumulates in certain tissues and organs, regroups and is subsequently excreted. The chemical composition of a radiopharmaceuticals determines its incorporation into kinetic or certain metabolic processes - targeted input (targeting) into relevant tissues, organs, cells or sub-cellular elements, including subsequent excretion. The built-in radionuclide then allows either external detection and imaging of the distribution of this substance (by gamma radiation in scintigraphy), or monitoring of its amount in samples taken (biological fluids, mostly blood or urine) - specific methods of these examination methods are described in detail below in the section "Clinical scintigraphic diagnostics in nuclear medicine".
   In the case of therapy, the radiation of the radionuclide performs biological effects on the cells of the tissue in which the radiopharmaceutical accumulates
(eg it destroys tumor cells - §6.6 "Radiotherapy", part "Radioisotope therapy").
   Radioindicators in nuclear medicine are applied in a small trace amount, about 10-9 -10-12 grams (pico- or nanomolar concentrations in tissues), so they themselves can not (bio)chemically affect the function of the examined organs, nor can they cause some side or toxic effects to the organism *). They can only cause radiation exposure, which we try to minimize by optimizing the applied activities.
*)
The only exception to this biochemical safety are radiopharmaceuticals based on murine monoclonal antibodies. In a small percentage of patients, they may experience allergic reactions due to the presence of so-called HAMA antibodies (discussed below in the section "Radionuclides and radiopharmaceuticals for scintigraphy").
   The best known example is the application of radioactive sodium iodide NaI131, which, like any iodine, is taken up (accumulated) in the thyroid gland. By external detection of gamma radiation emitted during radioactive b- transformations of 131I nuclei, it is then possible to measure the accumulation of this iodine or to display its distribution in the thyroid gland - §4.9.1 "Thyrological radioisotope diagnostics"; if desired, radiation b may have biological effects on the cells used in therapy, when higher activities are applied.
   A number of types of radiopharmaceuticals with affinity for the kidney, liver, bone, myocardium, some tumor or inflammatory tissues, signaling receptors have been developed, for the function of which the given substance is an indicator
(§4.8 "Radionuclides and radiopharmaceuticals for scintigraphy"). The degree of local accumulation of radiopharmaceuticals depends on the intensity of local metabolic and functional processes in organs and tissues. Using scintigraphic imaging, possible malfunctions can be located, analyzed and possibly quantified.
  Or the radionuclide is injected into the bloodstream and the dynamics of its passage trough the heart, lungs and large vessels is monitored - in this case without metabolic binding to a specific organ or tissue
(§4.9.4, part "Dynamic radiocardiography" and "Radionuclide gated ventriculography", or §4.9.8, part "Perfusion scintigraphy of the brain"); again with the possibility of analysis and quantification.
"Molecular imaging"
With the development of organic chemistry, biochemistry and cell biology, some radiopharmaceuticals have been developed whose labeled molecules have affinity for very specific cell types or processes at the subcellular level. With the help of scintigraphy and a suitable radiopharmaceutical, it is possible to purposefully examine not only the function of a certain organ or tissue, but also to selectively recognize a certain type of metabolic and transport pathway, such as enzyme or receptor binding or antigen-antibody reactions. For this purpose, special radiopharmaceuticals (both for diagnostics and for therapy) have been developed and are still being developed, which are characterized by their effects at the molecular level. With a bit of exaggeration, these methods of local measurement and imaging of the physiological response are referred to as "in vivo biochemistry".
Note:
 Of course, the name "molecular imaging" does not mean that we are imaging the molecules themselves (unfortunately, we cannot do that..), but we are depicting a distribution of the radioindicator that is a consequence and reflection of specific biochemical reactions at the molecular level.

Scintigraphy
The passage and distribution of a radioindicator thus reflects the specific physiological or pathological condition or function of the relevant organs and tissues. For its assessment in the simplest cases, it is sufficient to simply measure the intensity of radiation
g emanating from a certain place (eg from the thyroid gland - to determine its accumulation) by a collimated detection probe. For better and more comprehensive diagnostics, however, we need to measure - map out - display - the entire distribution of the radio indicator, including local details and anomalies. An important method called scintigraphy or gammagraphy is used for this :

Scintigraphy :
Scintigraphy or gammagraphy is a physical-electronic method of imaging the distribution of a radioindicator in an organism based on external detection of outgoing gamma radiation

Terminological note:
The more apt name of gammagraphy - gamma-ray imaging - is unfortunately used relatively rarely; predominant the less accurate name of scintigraphy, came from the fact, that scintillation detectors are now technically used here. In the future, scintillation detectors are likely to be replaced by semiconductor detectors
(see below "Alternative physical and technical principles of gamma cameras"), whereby the name "scintigraphy" already lost its justification. Out of inertia, however, name scintigraphy will undoubtedly persist.
  Scintigraphy or scintigraphic examination is often also called a scintigraphic study in the "jargon" of nuclear medicine. It dates back to the days, when scintigraphy was a new experimental research method to study physiological processes in the body.
   In most of the text of this chapter (§4.1-4.8) we will deal with the physical principles of scintigraphic imaging and technical solutions of devices for gammagraphic imaging. The clinical use of scintigraphy in nuclear medicine is summarized in the last §4.9 "Clinical scintigraphic diagnostics in nuclear medicine". And the therapeutic use of radionuclides is discussed in §3.6 "Radiotherapy", part "Radioisotope therapy".

Types of scintigraphy
Before we deal with specific physical-electronic methods for the implementation of scintigraphic imaging, we will briefly introduce the division (classification, categorization) of scintigraphic methods. In terms of time, scintigraphy can be divided into two types :

In terms of spatial-geometric, we can divide scintigraphy again into two categories :

In terms of complexity and interpretation of scintigraphic examination, we can distinguish two basic categories :

Radiation exposure at scintigraphic examination
At each interaction of ionizing radiation with an organism, some of this radiation is absorbed in the tissues and causes radiation exposure; at diagnostic applications (small) risk of unwanted stochastic effects. There is a significant difference between x-ray diagnostics and nuclear medicine in the laws of radiation exposure. During an X-ray examination, the source of ionizing radiation is the device (X-ray tube located outside the patient's body) and the radiation dose depends, among other things, on on the number of images taken, exposure times or the extent of the area scanned during CT
(§3.3, passage "Radiation load of patients during X-ray examination"). In scintigraphy, the source of radiation is not the diagnostic device, but the patient himself resp. radionuclide distributed inside his body in the investigated tissues and organs. We can then take any number of scintigraphic images, in different projections, with different acquisition times, without changing the patient's radiation exposure.
   The radiation dose received by the patient in connection with the scintigraphic examination is already given when the radiopharmaceutical is administered into the body. It depends mainly on the value of the applied activity [MBq] - direct proportionality. It also significantly depends on the type of applied radiopharmaceutical. The chemical form determines the degre and rate of accumulation of the radiopharmaceutical in various tissues and organs and the rate of its excretion. The radionuclide used for labeling determines the half-life of the radioactive transformation and the type of radiation emitted. In the case of pure gamma-radionuclides (such as
99mTc), the radiation burden is relatively low, since most of the penetrating g radiation passes through the tissue and carries its energy outward. However, with scintigraphy itself, the magnitude of the patient's radiation exposure does not depend on the acquisition time at all. The patient is continuously exposed to low levels of radiation (with a decreasing dose rate) even after leaving the nuclear medicine facility - during a subsequent stay at another healthcare facility or at home... The time during which radioactivity practically disappears from the body depends on the physical half-life of the radionuclide and the biological excretion half-life of the radiopharmaceutical; for radioindicators marked with 99mTc (approx 100-300 MBq), it usually takes about 2-3 days.
  In summary, issues of radiation exposure are discussed in §5.7 "
Radiation load during radiation diagnosis and therapy".

Basic principles of scintigraphic imaging
How to achieve gammagraphic imaging ?
 
An idea might arise to use photography for this: Radiation
g is an electromagnetic wave of the same physical nature as light. If we want to display an object using light (reflected or actively emitted), we use the laws of geometric optics and use a focusing lens to project an image of the object on a sensitive photographic layer and expose it for some time - a photochemical reaction creates a latent image, which after development becomes a visible image of different densities of silver grains in the photographic emulsion - see Fig.4.1.1 on the left.


Fig.4.1.1. Comparison of photographic imaging options in visible light and in gamma radiation.

It would be very pleasant if the patient could be "photographed" in this way in g radiation - Fig.4.1.1 in the middle. Unfortunately, this is not possible! Radiation g will not refract like light when it strikes the lens. As shown in §1.3, radiation g interact with each substance, and thus also with the material of the optical lens, in three ways :
1. Photoeffect- here the incoming photon ceases to exist and therefore will not arrive to the sensitive layer at all - it is not usable for imaging.
2.
Compton scattering - here would be some scattered photons g could strike the sensitive layer and elicit a photochemical reaction there, but the scattering angle is essentially random and always different, regardless of the angle of incidence. Compton-scattered radiation thus produces no image, but only a more or less monotonous graying or blackening of the film. Thus, even Compton scattering is not applicable for photographic imaging in gamma radiation *).
*) However, this statement is not completely absolute, it only applies to photographic images. At the end of §4.2 it will be shown that Compton scattering of radiation g can in principle be used for electronic collimation in so far experimental so-called Compton cameras .
3.
Formation of e- e+ -pairs (if the primary radiation g had energy >> 1MeV) - here the primary photon g disappears and the secondary photons of annihilation radiation always fly in opposite directions *), but each time at a different angle in space - unusable similarly to Compton scattering.
*) This property is used for electronic collimation in positron emission tomography (PET) - see §4.3, section "Positron emission tomography PET".
   We would reach the same conclusions if we tried to use a hollow mirror instead of a lens to display it in
g radiation. Only the simplest imaging using the pinhole camera in Fig.4.4.1 on the right, also works for gamma radiation, it is used in pinhole type collimators (they are described below in the section "Scintigraphic collimators").
   For radiation g therefore does not apply the laws of refraction and reflection => there is no refractive or reflective optics for radiation g ! We are not able to purposefully influence the direction of movement g -radiation photons *). Only for soft X-rays, under certain circumstances, reflective mirror optics partially work, but only for very small angles of incidence-reflection - see the appendix "X-ray telescopes" at the end of §3.2.
*) Physically conceived, only strong gravity can influence the direction of motion of photons
g (due to its universality). Although such gravitational lenses of gigantic dimensions are abundant in universe (see §4.3, passage "Gravitational lenses. Optics of black holes ." in the monograph "Gravity, black holes and space-time physics"), they are not feasible in laboratory conditions on Earth; even if we could make miniature black holes with the required properties, the quality of their images would not be very good and, most importantly, they would immediately kill us with their gravity and quantum radiation (§4.7 "Quantum radiation and thermodynamics of black holes" in the same book).
   The only way to achieve an imaging in g-radiation is collimation - shielding g radiation from all unwanted directions and releasing only radiation from the required direction. This creates a collimation projection in gamma radiation. In this way, most scintigraphy methods "works" with g radiation - see "Scintigraphic collimators" below. Exceptions are special methods using so-called electronic collimation by means of coincidence detection of two or more primary or secondary photons. These principles are used mainly in Positron emission tomography, or for so far experimental Compton cameras (see section "Compton cameras" and "High energy gamma cameras") or Compton telescopes in astrophysics - some "telescopes without lenses and mirrors"...

Motion scintigraph
Historically, the first type of instrument to perform scintigraphic imaging of the radioactivity distribution was a motion scintigraph, sometimes called a scanner. The first device of this kind was built in 1951 by B.Cassen and his colleagues, their main manufacturer in the 60s and 70s was the company Picker
(Fig.4.1.2 right). It is in principle a simple device, schematically shown in Fig.4.1.2 :


Fig.4.1.2. Motion scintigraph.
Left:
Principle diagram of the movement scintigraph
(bottom middle is an example of a thyroid scintigram) . Right: Scintigraph Picker 500i at KNM Ostrava.

A collimated scintillation detector *) is mounted at one end and an electromagnetic pen at the other end on a common massive arm moved by an electric motor. The detector shifts with a uniform meandering motion over the measurement object W, the radiation g (which is detected only from the area just below the collimator on its axis) is converted into electrical pulses, which (after amplification and amplitude discrimination, possibly reducing excessive frequency) are lead to electromagnetic coil. For each pulse, a ferromagnetic core is ejected from the solenoid coil, provided at the end with a pen (stamp), which prints a mark (comma) on the paper over the ink ribbon. Each comma represents, depending on the reduction setting, a hundred or a thousand pulses or the like. The higher the radioactivity of the place above which the collimated probe is located, the higher the frequency of pulses the probe will send to the solenoid coil and the denser the pen will type the commas of the image as it moves over the paper. The result is a display of the invisible distribution of the radioindicator using the visible density of commas on the paper (Fig.4.1.2 in the middle) - a scintigraphic image W* is created. In addition to paper, some instruments also recorded scintigrams on photographic film, which made it possible to better distinguish details in density of the commas - the frequency of pulses.
*) To increase the detection efficiency, relatively large scintillation crystals with a diameter of up to 15 cm, equipped with multi-hole focused collimators, were used. Thus, radiation g from the focus at the investigated site, from a relatively large spatial angle was concentrated on the surface of the crystal.
   The advantage of the motion scintigraph was simplicity and perhaps also the fact that it provided an image in a 1:1 scale. However, it had some major disadvantages. In the first place, it is a very low measurement efficiency: only a small part of the
g photons is always detected only from the place above which the detection probe is currently located - radiation from all other places escapes uselessly. Furthermore, the probe moves relatively slowly over the patient and takes a long time to scan the scintigraphic image. If the distribution of the radioindicator changes with time during the measurement, we are not able to capture and display these changes - the motion scintigraph does not allow dynamic scintigraphy. For these reasons, movement scintigraphs have not been used since about the end of the 1980s (they lasted the longest for thyroid scintigraphy, Fig. 4.2.1 in the middle of the bottom) - then they were completely replaced by scintillation gamma cameras.


4.2. Scintillation gamma cameras

A scintillation camera is a device that detects photons of radiation g simultaneously from the entire field of view, converts them into electrical impulses and then uses them to create a scintigraphic image of the distribution of the radioindicator in this field of view.

The principle of the scintillation camera
Scintillation cameras, or gamma cameras, are so far the most perfect devices for scintigraphic imaging of radioactivity distribution. It is a very complex device both in its principle and in its technical construction.
  The first scintillation camera was constructed by H.O.Anger in 1958. In the initial experiments, he used a single-hole collimator and the scintillation in a thin crystal of larger diameter exposed to a photographic plate. He achieved a striking improvement by attaching photomultipliers (originally 7 photomultipliers) to the crystal, which sensed flashes in the scintillation crystal and converted them into electrical pulses that were electronically evaluated. The first scintillation cameras with 19 photomultipliers began producing company Nuclear Chicago in 1964, soon to be Picker (a leading manufacturer of motion scintigraphs); later in Europe Intertechnique, Philips, Gamma, in Japan Toshiba.
   The schematic diagram of Anger's scintillation camera is shown in Fig.4.2.1 :


Fig.4.2.1. Schematic diagram of a scintillation camera (analog).
Note: For clarity, only two photomultipliers F1 and F2 are shown. In fact, there are a larger number of photomultipliers - min. 19 (for older cameras with a smaller field of view), 32, 64 and more.

Detection of radiation g and determination of the place of its origin
Let's consider a (model) investigated object
W, in which there are three localized deposits A, B, C of increased concentration of g- radioindicator. From each place of deposition of radioactivity, radiation g is emitted isotropically on all sides, which, due to its penetration, emanates from the object W out. In order for this radiation g to be able to create an image, a collimation projection must first be performed. We achieve this by putting a lead plate in the path of the emitted radiation g, drilled with a large number of small parallel holes. Only those photons g, that move exactly in the direction of the axis of the holes, can pass through this collimator. Other photons that go "obliquely" are absorbed on the lead partitions between the holes. The collimator thus creates a planar projection of the radio indicator distribution into the blue marked plane in Fig.4.2.1. A thin large-area scintillation crystal is placed here. Each photon of radiation g that passes through the collimator causes a scintillation flash of a large number of photons of (visible) light in the crystal. Scintillations from crystal are sensed and converted into electrical pulses by a system of photomultipliers, optically adhered to the crystal *). For simplicity, only two photomultipliers are drawn in Fig.4.2.1 - F1 and F2.
*) The general principle of scintillation detectors and photomultipliers, their properties and construction are discussed in detail in §2.4 "Scintilltion detectors".
   Let us now observe the "fate" of the individual photons 
g emitted from the interior of the object W under investigation. In particular, any photon g' that flies in a direction other than exactly perpendicular to the collimator face (i.e., parallel to the orifice axes) is absorbed at the partitions between the collimator orifices, does not fall on the crystal, and is not detected. The photon gA, which flies in the "right direction" from position A, passes through the collimator opening and causes at position in the crystal a scintillation, whose photons propagate in all directions in the crystal. A photomultiplier F1, which is close to the site of scintillation, will receive a relatively large number of photons from this flash, so that the pulse at its output will have a high amplitude, while the distant photomultiplier F2 will receive only a small portion of these photons and its pulse will be very low. For the photon gB from position B, scintillation occurs approximately midway between the photomultipliers F1 and F2, so that the amplitude of their pulses will be approximately the same. For photon gC (radiated from deposit C), which impact to the crystal and causes scintillation near the photomultiplier F2, the photomultiplier F2 will receive much more light than the photomultiplier F1, and this will also be the ratio of the amplitude of their pulses.
   In general, most light enters the photomultiplier, which is closest *) to the flash point
(the point of interaction of the photon g with the crystal) - therefore a pulse is generated at its output, the amplitude of which is larger than the amplitude of pulses from more distant photomultipliers, whose phocathodes receive less light from a given flash. The localization of the flash positions is thus performed by a kind of electronic-geometric "triangulation", is determined as the "center of gravity" of the signals from the photomultipliers.
*) The photomultiplier receives the largest portion of light when scintillation occurs directly below the center of the photocathode. From scintillations at more distant locations, fewer photons will fall on the photocathode, so the output signal has a lower amplitude.
  Thus, we see that by comparing the amplitudes of the pulses from the individual photomultipliers, it is possible to calculate the position of the scintillation in the crystal, and thus the place in the patient's body, from which the photon g was emitted. Pulses from individual photomultipliers (of which there are a larger number - e.g.16 (for older cameras with a smaller crystal), 32, 64 and more), are led to an electrical circuit called a comparator (based on a resistive matrix), where the pulse amplitudes are compared and the resulting X and Y coordinate pulses are generated - these already carry direct information about the position of scintillation in the crystal, and thus also about the position of the place in the organism from which the respective gamma photon was emitted. After amplification, these X and Y pulses are fed to the deflection plates of the oscilloscope screen, where they determine the position of the flash on the screen (this was the case with older analog gamma cameras used in the 1960s and 1970s).
Amplitude analyzer
In addition to coordinate analysis, pulses from all photomultipliers are fed to the summing circuit - from the point of view of this circuit, the whole scintillation camera behaves as one large scintillation detector of radiation
g. These summation pulses, the amplitude of which is proportional to the energy of the absorbed radiation g, are then sent to an amplitude analyzer *) (pulse selector according to amplitude) - for each flash is thus determined not only its position (coordinate pulses X, Y), but also the energy of the photon g, which this flash caused. The analyzer window is set so as to transmit only pulses corresponding to the photopeak - total absorption of radiation g in the crystal. If the radionuclide used has more radiation energies g, the window is usually set to the "main" (strongest) photopeak, or measurements in multiple windows set to individual photopeaks shall be used.
*) The principle and role of the amplitude analyzer in radiation spectrometry is described in §2.4 "Scintillation detectors".
   For correct radiometric measurements on each spectrometric instrument, the basic condition is to set the analyzer window to the photopeak of the gamma radiation of the used radionuclide. In the case of a scintillation camera, in addition to the detection efficiency, the correct adjustment of the analyzer window is necessary to suppress Compton scattered radiation and to ensure the alignment of the photomultipliers to achieve good field of view homogeneity (see below passage "Adverse effects with scintigraphy and their correction", part "Compton scattering g).
   In older types of gamma cameras, the analyzer window was set to photopeak manually, with modern digital cameras is implemented automatic setup and tuning window analyzer - called Peaking or Auto Peak (automatic tuning peak). By comparing the frequency of pulses in the lower and upper half of the window, this window analyzer is automatically tunes to the center of photopeak (see figure) :
                            
Formation of analog scintigraphic image
The pulses behind the amplitude analyzer, called Z
(have nothing to do with the third dimension coordinate!) are uniform "trigger pulses" - they say: "Yes, a 'correct' g photon has now been registered and the X and Y coordinate pulses are valid". The Z pulses are fed to the grid of the oscilloscope screen; here it cancels the negative bias for a moment, causing the cloud of electrons to emerge from the cathode, focusing and accelerating in the "electron cannon" and flying towards the screen screen. In the meantime, the X and Y coordinate pulses have already appeared on the accelerating plates, whereby the electron beam is deflected in the appropriate direction and falls into the appropriate place (A*, B*, C* - depending on point where the photon g is emitted - A, B or C) of the oscilloscope screen, where it emits a flash of light. As the flashes gradually come to the screen as if they were "raining" there, these analog images are sometimes called "images with rain".
   In this way, the invisible distribution of the radioindicator in the examined object W, via physical-electronic detection of invisible gamma radiation, is displayed in the form of a density of visible flashes in the corresponding places of the screen - a scintigraphic image W* is created. Radioactive structures (lesions) A, B, C in the examined object are displayed as sites A*, B*, C* with increased number of flashes on the screen.

  The described scintillation camera according to Fig.4.2.1 provides analog scintigraphic images on the oscilloscope screen. This image is present here only for the duration of the photon g scan by the gamma camera, after end of the scanning ("patient departure") this image disappears. To preserve this image, it was photographed from the screen with a camera, whose shutter was open while the pulses were being stored. The so-called persistent oscilloscope was also often used, on the screen of which the flashes did not disappear immediately, but remained here for an adjustable time and only then gradually faded until they disappeared.

Digital scintigraphic images
The above-described photographic method of recording (analog) scintigraphic images has the disadvantage, that it cannot be post-edited (need intensifying dark underexposed areas and weakening bright overexposed places) and, most importantly, it cannot be quantified. Therefore, with the development of desktop minicomputers in the 1960s, there was an effort to supplement (and later replace) oscilloscopic imaging of analog scintigraphic images by digitizing them and storing this images into the computer memory. The scheme of operation of such a gamma camera equipped with an acquisition computer is shown in Fig.4.2.3 :


Fig.4.2.3. Creation of a digital scintigraphic image by AD-conversion of analog X, Y coordinate pulses, their storage in the image matrix of the computer memory and display on the monitor screen.

The scintillation camera itself and the relevant electronic circuits for amplification, comparison, summation and amplitude analysis of pulses are identical as in Fig.4.2.2. Only the oscilloscope screen in the right part is replaced by a special circuit - the so-called analog-to-digital converter ADC (Analog-to- Digital Converter) and computer memory. The actual conversion process is started by the trigger pulse Z, which indicates that a valid photon of radiation g has been detected. The amplitudes of the X and Y coordinate pulses are then converted by the ADC converter into digital (numerical) information - a bit combination - and sent to the corresponding cell address in the computer. A certain sequence of cells is set aside in the computer's memory to write these digitized pulses; these cells are software-arranged into a so-called image matrix - it is usually 64x64, 128x128, 256x256 cells (exceptionally also 512x512 cells; for cameras with a rectangular field, then neither the image matrix is not square). Each cell in the image matrix topographically corresponding to a specific location in the displayed object W . The field of view of the gamma camera is thus divided by a grid into small squares - pixels (picture element), which correspond to individual addresses in a defined part of the acquisition computer's memory.
   Before the start of the acquisition, the contents of all cells are reset. If a digitized pulse arrives at a cell from the ADC converter, its content is increased by 1. Thus, photons of radiation
g, converted into electrical pulses and digitized, gradually populate the cells in the image matrix of the computer memory, according to the place of the radiation emission, with ever-increasing values of their content - a digital scintigraphic image formed by the numerical content of the image matrix cells in the computer memory. The numerical content of each of these memory cells (pixels) is directly proportional to the radioactivity corresponding site in the organism, resp. its columnar projections from the entire depth of the displayed area. The image matrix from the computer's memory is then electronically displayed ("mapped") on the computer monitor screen.
FRAME mode, LIST mode 
The above described method of cumulative explicit recording the scintigraphic image into memory is called a frame mode
("image method "). For special purposes (for phase dynamic studies and iterative tomographic methods - §4.3, part "Computer reconstruction of SPECT", "Reconstruction of PET images ", "TOF - time localization of the annihilation site") is sometimes used so called list mode ("list method "), where only a list of X and Y coordinate values of successive incoming pulses (together with time stamps) is sequentially loaded into memory and the own images are created additionally only after the acquisition is completed.
Digital scintillation cameras 
With the development of electronics, especially the construction of fast and miniaturized ADC-converters and microprocessors, the digitization of the scintigraphic signal is no longer limited to the conversion of "finished" analog X, Y coordinate pulses according to Fig.4.2.3. With current so-called digital gamma cameras, each photomultiplier already has its own analog-to-digital ADC converter at its output. The calculation of the coordinates of scintillation in the crystal is not performed in an analog comparator, but in a digital microprocessor, which already directly "populates" the respective addresses in the computer's image matrix with the relevant numerical information. In addition, the gain of the preamplifier of each photomultiplier via a DAC converter is controlled directly from the computer, which allows more accurate and operative calibration of the camera - adjustment (tuning) and setting of appropriate corrections for homogeneity and linearity.

Construction arrangement of scintillation cameras
Gamma camera detector 
A large-area scintillation crystal of a gamma camera with glued photomultipliers (their number is usually 19 to about 120) and appropriate electronics is built into a special robust housing (a kind of "pot"), providing light tightness and radiation shielding against ambient ionizing radiation. The metal housing also shields the photomultipliers against an external magnetic field. At the bottom of the camera housing is a mechanism for attachment the collimator, which must be tightly attached to the crystal. The collimators are exchangeable, during manual exchange they are usually fastened with screws, for automatic exchange the collimators are fixed with special motorized holders. For SPECT cameras, there are also touch sensors for mechanical protection of the patient and the detector when the camera moves towards the patient.
Stand and gantry for mounting detectors 
The entire camera detector is then mounted on a special stand equipped with electric motors for mechanical movement of the camera - shift in the vertical, or event. horizontal direction and for rotation of the detector. For SPECT tomographic cameras, the stand is made in an annular arrangement as a so-called gantry, enabling by use of an electric motor angular rotation of the camera around the examined object. There are usually two detectors mounted on the gantry, which can be angulary rotated around the axis of the lounger - a "double-headed" camera. Additional electric motors ensure radial displacement of the detectors towards the center and away from the center, so that it is always possible to set the smallest possible distance between the body surface and the collimator face.
Examination lounger 
Under the camera detector, there is a bed (lounger) for the examined patient - perpendicular to the stand, or enters inside the gantry. Manually or motorized, it allows horizontal movement in a sufficiently large range
(up to 2m) to be able to pass with the whole patient under the camera or through the gantry and take images of different parts of the body. To a lesser extent (approx. 60 cm) a vertical shift is also realized. The lounger should be sufficiently robust (load capacity min. 180 kg) and stable, ensuring mechanical positioning with the possibility of locking. The support plate of the lounger in SPECT cameras is made of a material with low absorption of gamma and X-rays (when scanning from the front and back through the lounger). With the lounger pushed aside and the camera detector turned vertically, scintigraphic examinations of patients can also be performed sitting or standing.
   To perform the whole body scintigraphy (whole-body imaging), the bed with the patient with using the electro-motor is slowly moved in the longitudinal direction, so that the individual parts of the patient's body gradually enter the field of view and are detected by the camera detectors; the acquisition computer fluently composes of a whole-body scintigraphic image - "gliding" whole-body scintigraphy.

Auto-Body-contouring
 
To achieve the best possible resolution, the gamma camera (collimator face) should be placed as close as possible to the patient's body surface
(trigonometric analysis is performed below in §4.5, section "Spatial resolution"). Auto-contouring or body-contouring is a useful opto-electronic tool for ensuring optimal quality of scintigraphic imaging in whole-body and SPECT examinations: when moving the lounger and rotating the camera, using electronic position sensors, the camera detectors on the gantry are automatically shifted by electric motors so that they "copy" the patient's body and the collimator is still as close as possible to the patient's body surface (automatic "body contouring").
  Auto-contouring is realized by means of two rows of infrared LED diodes and two rows of opposite photodiodes, placed in two strips mounted on opposite edges of the camera detectors. Electronic circuits regulate the radial position of the gamma cameras so that the infrared rays from the outer row are interrupted, but not from the inner row (closer to the front of the collimator). The distance of the detector is thus constantly kept in the range between the two rows of LEDs <--> photodiodes, approx 10 mm.


Fig.4.2.4. Construction arrangement of a scintillation camera.
Left: Uncovered scintillation camera detector - collimator, crystal, system of photomultipliers and electronic circuits.
Right: Example of an assembled planar camera with one detector (top) and a SPECT tomographic camera with two detectors in gantry (bottom).

In the left part of Fig.4.2.4 is a disassembled detector of a smaller older camera (PhoGamma Nuclear Chicago, with 19 photomultipliers), removed from the shielding package. Below we see a collimator, above it is a thin circular scintillation crystal, to which photomultipliers are optically attached via light guide blocks. In the upper part of the detector there is the appropriate electronics, especially the preamplifier for each photomultiplier, adjustment circuits, for digital cameras also analog-to-digital converters and microprocessors for determining coordinate pulses. Newer scintillation cameras have a larger rectangular crystal, equipped with a larger number of photomultipliers.
   In the right part of Fig.4.2.4 there is an example of two installed cameras. Above is a smaller planar camera with one detector on a simple stand
(PhoGamma HP from 1973, with Clincom evaluation device; on the left next to the camera stand, there is a stand with interchangeable collimators), at the bottom there is a larger SPECT tomographic camera (from 2002) with two detectors ("heads") mounted on circular gantry *) and motorized movement of a lounger for whole-body scintigraphy.
*) Occasionally was also used some other construction arrangements of scintillation camera devices (Anger-type camera detectors themselves are designed almost identically for different types and manufacturers; other alternative technical solutions are mentioned below). Instead of the classic circular gantry, the detectors were mounted on special arms, the movements of which were electronically controlled by servomotors. The advantage here was perhaps greater flexibility of different detector positions (including the possibility of simultaneous independent sensing of two patients by each detector separately). In addition to "universal" cameras, special single-purpose cameras with a fixed detector configuration were sometimes used, such as 3 or 4 detectors connected in a triangle or square, designed for scintigraphy of the heart (myocardium) or brain. However, all these more complex construction arrangements of gamma cameras did not not proven himself in the end and soon ceased to be used...
   The electronic circuits of the scintillation camera have been described above
(in the section " Principle of the scintillation camera ") only in a general and simplified way, rather from a physical point of view. Scintillation cameras are equipped with a number of other electronic circuits for adjustments and for corrections of physical-electronic influences. They are important eg circuits for the correction of X, Y coordinate pulses - the shape and size of the image, especially the correction of the dependence of the image size on the energy of the detected gamma radiation - so that the scale of the image is not dependent on this energy.

Scintigraphic collimators
The primary "optical member" of a scintillation camera, through which radiation
g is the first to pass, is the collimator *). In terms of gamma imaging, the collimator has an analogous function as an optical lens when photography. Its task is to make the most perfect projection of the distribution of radioactivity in the examined object using g- radiation into the plane of a large-area scintillation crystal. Therefore, the final quality of the scintigraphic image largely depends on the properties of the collimator.
*) From the general point of view of radiation physics and radiation detection, collimators were discussed in §2.1 "
Methodology of ionizing radiation detection", paragraph "Shielding, collimation and filtration of detected radiation" and in §3.1 "Nuclear and radiation methods", section "Collimation of ionizing radiation"). In scintigraphy, collimators have an imaging role. For positron emission tomography, coincident electronic collimation is used instead of mechanical collimators for imaging - see below “Positron emission tomography PET”.
   In general, the collimator is a special aperture made of a shielding material (mostly lead, sometimes tungsten), defining the direction of the photons incident on the scintillation crystal as well as the field of view of the camera. Most often it is a plate with a large number of densely and evenly spaced holes - channels - of a certain shape, size and direction. Without attenuation, only photons flying in the direction of the axis of the collimator's orifices pass through the collimator (and impinge on the crystal), or only with a small deviation, ie almost perpendicular to the collimator front and to the crystal surface. Other photons in other directions are absorbed in lead partitions (septs, baffles) between the holes, they do not fall on the crystal and are not detected.

   Collimators for scintillation cameras are usually replaceable - there are several types of collimators with unambiguously defined properties, which govern their use. The collimators are distinguished according to the number, size and configuration of the holes, according to the radiation energy g for which they are optimized, according to the resolution and sensitivity (detection efficiency). The imaging properties of collimators are discussed in more detail in §4.5 "Physical parameters of scintigraphy".
   Here we give a brief overview of the basic types of collimators - Fig.4.2.6. First we will deal with collimators with parallel holes - channels - perpendicular to the scintillation crystal of the camera, which are by far the most common type - here the image of the object created in the detector has the same size of 1:1 as the displayed object, regardless of the distance of the source from the collimator
(however, the spatial resolution of the imaging depends significantly on this distance, see below).


Fig.4.2.6. Left: Basic types of collimators of scintillation camera (gamma camera crystal is in the up position, just above the collimator). Right: Example of a robust high energy collimator HE and a subtle low energy collimator LE HR
(and cutout from UHR) .

Collimators for different energies
The most basic criterion according to which collimators are divided is the radiation energy
g, for whose scintigraphic imaging the collimators are optimized. According to this gamma radiation energy, the collimators have different thicknesses of the partitions (septums) between the openings *), sufficient to absorb the radiation of a given energy.
*)
The thickness of the partitions 
The optimization of the collimator design for the required energy of gamma photons is based on the requirement, that gamma radiation passes only through the holes, while in the partitions (septs) between them it was effectively absorbed. If gamma radiation penetrated to a greater extent across the baffles, it would degrade the imaging properties of the collimator, especially the contrast
of the image (it is discussed in §4.5, passage "Over-radiating trough collimator septa", Fig.4.5.3). For complete absorption of gamma photons would need a large thickness of the baffles, which would lead to very low detection efficiency. However, as a sufficient criterion for achieving a reasonable level of cross-radiation over baffles, without significant deterioration of the image contrast, a value of 5% is considered. According to the trigonometric analysis in Fig.4.5.3b in the passage "Over-radiating trough collimator septa", this leads to the condition for the transmission factor e -m .s.L /(2d + s) <0.05, where d is the diameter of the holes, L their length, s the thickness of the baffles and m is the linear absorption coefficient of the collimator material (lead) for the required gamma energy. This gives a limitation for the thickness of the collimator septs s > (6.d/m)/[L - (3/m)]. The optimal is the smallest possible thickness of the partitions, allowed by cross-radiation - so that the septa shades the smallest possible area of the detector and the efficiency (luminosity) of the collimator is the best possible.
  The absorption coefficient of the collimator material (lead) strongly depends on the gamma energy, on which thus the required thickness of the baffles depends. For low energies around 150keV, where for lead is
m » 21.4 cm-1, eg for a collimator with holes 2 mm in diameter and 25 mm long, the required partitions thickness is s » 0.3 mm (thin lead foil). For higher energies around 400keV, where m is » 2.5 cm-1, significantly thicker partitions s » 4.5 mm are needed.
   According to gamma energy we have 4 basic types of collimators (Fig.4.2.6 left) :

Recently, it has been constructed :

Appropriate selection of the collimator according to the energy of the emitted gamma radiation has a fundamental effect on the quality of the scintigraphic image. For low energies, such as 140keV 99mTc, we use Low Energy collimators, which provide the best resolution. If we used a robust HE collimator (for high energies) here, we would get an image with lower resolution and lower detection efficiency, on which, in addition, the lead septa between the holes of the collimator *) would be disturbingly visible. We can also use the Pinhole collimator (see below "Collimators with special geometry"), which provides a quality image, but with significantly lower detection efficiency. For higher energies, such as 364 keV 131I, the collimators Low Energy are completely unusable, significant cross-radiation between the septa completely degrades the image into a shapeless "daub" (it is discussed in §4.5, passage "Cross-radiation of the collimator septa"). It is imperative that we use a High Energy collimator here (the holes and partitions of the collimator can also be seen in the picture) or Pinhole. Pinhole is the only type of collimator, that is in a wide range independent of energy.
*) This disturbing structure of the holes and septa of the HE collimator can be suppressed by a stronger smoothing of the image (approx. 4
x S9), at the cost of a lower resolution - pictures on the right.


Scintigraphic images of a thyroid phantom filled with
99mTc (top) and 131I (bottom), imaged using the collimators Pinhole , Low Energy HR and High Energy HE. The disturbing display of the holes and septa of the HE collimator can be suppressed by a stronger smoothing (filtering) of the image - pictures on the right.

Collimators according to resolution and sensitivity
Another criterion for the division of collimators is their required resolution and sensitivity (efficiency - "luminosity"). However, this only applies to low energy LE collimators; with robust collimators for high and medium energies we cannot achieve either good resolution nor high sensitivity, due to the thick partitions between the holes
(and thus the low density of the holes). According to the resolution and sensitivity, low-energy collimators are further divided into :

The number of collimator holes
depends on the type of collimator and its size (area) of the camera's field of view. With the current planar/SPECT cameras, the field of view is around 55
x 45 cm. The total number of holes for the basic types of collimators is then approximately :
HE - 8000 holes ; ME - 15,000 holes ; LEAP - 80,000 holes ; LE HR (UHR) - 140,000 holes .
The holes are usually hexagonal in shape.

Spatial resolution of a gamma camera
The spatial resolution of a camera is determined by two components: the internal resolution of the detector and the resolution of the collimator
(for a more detailed analysis, see §4.5, section "Spatial resolution") . The resolution of the collimator is determined by the diameter of the holes and their length. HR collimators with narrow and long holes (the length of the holes is given by the thickness of the collimator) have better resolution than thinner HS collimators with larger and shorter holes. The spatial resolution of the gamma camera significantly depends on the distance displayed structures from the collimator front. From each hole of the parallel collimator we can draw an imaginary cone defining the area from which gamma radiation can pass through this hole to the camera detector (radiation from places outside this cone is absorbed by the lead septa of the collimator). With the distance from the collimator, this detection cone widens, which significantly worsens the geometric spatial resolution of the image projected by the collimator on the scintillation crystal of the gamma camera (trigonometric analysis is performed below in §4.5, section "Spatial resolution", here for the sake of clarity we present only the basic Fig. 4.5.2 :) .


Fig.4.5.2. Deterioration of the positional resolution of the gamma camera with increasing distance h from the collimator front. The image of the point source becomes more and more "blurred" with increasing distance, the PSF expands and the spatial resolution of the FWHM deteriorates - Fig. d). Deterioration of the spatial resolution is accompanied by a decrease in the brightness of the image, but the total number of pulses is the same in all images and the area (integral) under the PSF function is also the same for all distances.

The gamma camera (front of the collimator) should therefore be placed as close as possible to the surface of the patient's body. For collimators with a different arrangement of holes (see below), the geometric situation is more complicated, but in principle the same rule applies to the deterioration of the spatial resolution for greater distances from the collimator face.
Detection efficiency of the scintillation camera 
The detection efficiency (sensitivity) of the camera is given by the efficiency (luminosity) of the collimator and the internal detection efficiency of the detector (discussed in more detail in §4.5, section "Detection efficiency (sensitivity) of the gamma camera"). Efficiency (transmittance, luminosity) of the collimator is given by the diameter of the holes and their length, but in the opposite ratio to the resolution. The larger and shorter the holes, the higher the detection efficiency. The efficiency or luminosity of collimators is generally very low - around 1-2 %.
   Interestingly, with gamma cameras, when using parallel collimators, the detection efficiency (sensitivity) does not depend on the distance h of the displayed source from the collimator front! The imaging of the point source in a wide range of distances 0-30 cm from the front of the collimator in Fig.4.5.2 d shows a deterioration of spatial resolution and decreased image brightness, but the total number of pulses is the same in all images, area
(integral) under the PSF function is the same for all distances. This surprising behavior is due to the specific properties of geometric collimation in parallel collimators. We can clearly illustrate this according to the schematic drawing in Fig.4.5.2 b) as follows: As the source moves away from the collimator front, the number of photons incident on the individual holes decreases quadratically as 1/h2. However, the number of holes through which radiation can pass to the detector, increases quadratically in proportion to h2. These two opposing trends cancel each other out, so the total photon flux passing through - collimator efficiency - does not change with the distance between the source and the collimator.
Note: This rule does not apply to special convergent or Pinhole collimators, the detection efficiency here changes significantly with distance - it increases or decreases (see §4.5, section "Imaging properties of special collimators").
   However, this distance sensitivity independence of parallel collimators only applies to situations without a substance-absorbing environment - in vacuum or in air. In practical scintigraphy, however, there is a tissue environment between the displayed structures with distributed radioactivity in the organism and the gamma camera, with which gamma radiation interacts, which leads to the absorption and attenuation of gamma radiation. This a gamma-ray absorption, also called attenuation, is reflected in scintigraphic images by an artificial reduction in the number of pulses from structures deposited at greater depths, compared to structures closer to the surface. In such a case, the statement that the detection efficiency (sensitivity) does not depend on the distance of the displayed source from the collimator front, is no longer valid. Here, the detection efficiency decreases significantly with the distance - depth - of the displayed source !

Collimators with special geometries
In addition to collimators with parallel holes - channels - the collimators with otherwise geometrically arranged holes are used for some special purposes (Fig.4.2.6. in the middle) :

The imaging properties of collimators are discussed in more detail in §4.5 "Physical parameters of scintigraphy". Here, for clarity, we will only duplicate graphs of the dependence of the spatial resolution and detection efficiency (sensitivity) of the gamma camera with basic collimators on the distance :

Fig.4.5.6. Dependences of the spatial resolution FWHM (left) and the detection efficiency S (right) of the gamma camera on the distance of the source from the front of various types of collimators.
   Imaging properties of the most important types of collimators with different geometric arrangement of the holes, we tested using linear orthogonal grid (its construction is described in "Phantoms and phantom measurements in nuclear medicine" image "Grid") :

For a collimator with parallel holes (such as LE HR left) we get a linear imaging of the grid everywhwre, only for a greater distance from the front of the collimator, the spatial resolution deteriorates (blured grid). With a convergent collimator (such as a SmartZoom with the convergent center part) the image of the center part increases with increasing distance. With the Fan Beam collimator (which is convergent in the transverse direction, parallel in the axial direction), the grating espands only in the transverse direction with increasing distance, it remains the same in the axial direction.
   The most striking dependence on the object distance exhibits the collimator Pinhole: tightly close to the opening we get the image magnified many times, with increasing distance the zoom decreases and for distances above approx. 20cm the image is already reduced.
   Of all the images is also seen a general trend of deteriorating resolution (and thus contrast in the image) with the distance from the front of the collimator.

Scintigraphic images and their evaluation
The whole process of scintigraphic diagnostics is schematically shown in Fig.4.2.5. After application of radioindicator, its distribution occurs in certain parts of the organism
(uptake in target tissues and organs, or flow of the radiotracer trough blood vessels and heart cavities). This distribution is imaged by a scintillation camera using external detection of the emitted radiation g. Digital scintigraphic images are created on a computer, which on the one hand we evaluate visually, or we can create curves and mathematically analyze the investigated processes and calculate quantitative parameters of the functions of individual organs. Finally, an interpretation of all these partial data and results is coming, which, together with data from other methods, will result in the making a diagnosis in the final protocol.


Fig.4.2.5. Schematic representation of the whole process of scintigraphic examination - from the application of the radioindicator to the patient and its uptake in target tissues and organs, through the process of scintigraphic imaging with a gamma camera, visual evaluation of images, mathematical analysis and quantification, to interpretation and making a diagnosis.

The methodology of mathematical analysis and computer evaluation of scintigraphic studies will be discussed in more detail below in Chapter 4.7 "Mathematical Analysis and Computer Evaluation in Nuclear Medicine".

Adverse influences on scintigraphy and their correction
In scintigraphy, there are some unfavorable and disturbing phenomena, which can worsen the quality of the image and thus, in the extreme case, even lead to incorrect interpretation of scintigraphic examinations in the sense of false negative or false positive findings. Here are six basic adverse effects that occur in general in every scintigraphy, ie in planar scintigraphy and SPECT tomographic scintigraphy. Other adverse and disturbing phenomena specific to SPECT (such as instability of the axis of rotation or artifacts arising during reconstruction) and PET (random false coincidences) will be mentioned below in §4.3.


Fig.4.2.8. Influence of registered number of photons on image quality in terms of statistical fluctuations (noise) - image quality improves with increasing number of photons.
Above: Photographic portrait exposed with different number of photons of light .
Bottom: Gammagraphic image of a phantom
(Jasczak, filled with 99mTc radionuclide ) taken by a scintillation camera with different numbers of g- photons in the image.

Physical parameters of scintigraphy
Resolution, detection efficiency, homogeneity and other parameters of the scintillation camera are defined and discussed below in §4.5 "
Physical parameters of scintigraphy - image quality and phantom measurements". The methods of their measurement and testing are discussed in the work "Phantoms and phantom measurements in nuclear medicine".

Errors and pitfalls of correction methods - correction artifacts
It should be noted that no correction methods are "self-saving", but they can have their pitfalls. Errors of correction methods can be divided into two categories :

  Undercorrection, overcorrection, and correction artifacts can lead to similar deterioration (or even the risk of misinterpretation) of scintigraphic images as uncorrected studies. Experience shows, that in order to correctly interpret the findings, it is necessary to carefully compare images without correction and images with correction by the "trained eye" of an erudite expert, who must also take into account the specific anatomical and positional circumstances of the patient.

Scintigraphic image quality - lesion recognition
The above-mentioned adverse effects mean that the scintigraphic image is not entirely accurate and perfect - despite useful information, disturbing statistical fluctuations (noise) overlap, the image is blured and often low in contrast. This imperfect quality leads to the fact that some more subtle structures of the examined object are not visible on the scintigraphic image - we say that such lesions are not detectable. In the diagnostic practice of nuclear medicine, such a scintigraphic image is optimal, which, in addition to objectively measurable physical parameters, it also suits the human subjective visual perception of the evaluating physician. So what parameters of the examined object and its image decide on the the objective imaging and the best possible recognition of lesions ?
  The basic regularities result from the properties of scintigraphic imaging and from the statistical analysis of the resulting image data. In the left part obr.4.2.9 shows scintigraphic images of simple structure (lesion) of the circular shape of the size (diameter) d and the specific activity A, surrounded by a homogeneous environment - background - specific activity of B . The prominence of the lesion against the background can be characterized as the contrast of the object C
obj = (A - B)/B; (event. x 100 in [%]). Scintigraphic imaging produces an image in which the lesion is shown as a structure A* and the background as a constant area B* (more or less wavy,depending on statistical fluctuations).


Fig.4.2.9. Analysis of contrast and statistical fluctuations of scintigraphic imaging of lesions
(phantom measurements on a PhoGamma LFOV camera) .

If we compare the original object with its scintigraphic image, we see two main differences :
¨ 1. Blur and contrast reduction   
Due to imperfect spatial resolution, the sharp contours of the original object A were blurred and the difference between image maximum A* and background B* decreased - image contrast C
img = (A*max - B*)/B* is lower than the contrast of the object Cobj : C img < C obj . Assuming a circular lesion and Gaussian convolutional blur (the response function of the point source PSF of the camera has the shape of a Gaussian curve with a half-width FWHM), the relationship between the contrast of the object and the image is given by the exponential expression :
                    C
img  =  C obj . e - (FWHM / d) 2  ,
where FWHM is the camera resolution and d is the size (diameter) of the lesion. For large lesions (d> 4.FWHM), the contrast of the image hardly changes (C
img @ Cobj). However, in small lesions, comparable or smaller than the FWHM camera resolution, the contrast degradation is very significant, Cimg << Cobj (at a typical camera resolution of 10mm, the contrast of a 1cm lesion decreases almost 3-fold, in a 5mm lesion more than 30-fold !).
¨2. Statistical fluctuations - noise 
Due to the quantum stochastic laws of radioactive decay, emission and detection of quantum radiation
g, all parts of the scintigraphic image show statistical fluctuations - noise is covered over the image of the object. As shown in §2.11 "Statistical fluctuations and measurement errors", the magnitude of this noise at each point of the image is given by the square root of the average accumulated number of pulses n : s = ±Ö(n). The relative statistical fluctuations s/n = 1/Ö(n) are lower the higher the number of pulses accumulated in the individual cells of the image. Constant background B is thus shown as an area whose points fluctuate roughly between B* ± Ö(B*), ie with sB = ±Ö(B*). Similarly, the point values in the A* image fluctuate statistically. If these fluctuations are too high, comparable to the average values of the difference between A* and B*, these differences can easily be "lost" in them and the corresponding structure will not be visible in the image. Disturbing statistical fluctuations are thus a fundamental limiting factor for the recognizability *) of small and not very contrasting lesions on the scintigraphic image.
*) Were if not for statistical fluctuations, by artificial increase in steepness (contrast) display of a scintigraphic image on the screen, to would be possible achieve visibility of even small and low contrast lesions. In addition, appropriate deconvolution filtering (using the inverse modulation transfer function MTF) could correct the camera resolution, resolution recovery - computer "focus" of the image - and reconstruct all details from the displayed object (see "Filters and filtering", section "Band focus filters"). Unfortunately, statistical fluctuations deprive us of most of these possibilities in practice ...
   The statistical analysis of image data shows, that we can only recognize (and statistically prove) in the image a structure (lesion) whose contrast C
img satisfies condition
                    C
img  >  4 / Ö(B*)  .
It is a condition of the statistical significance of the difference A*
- B* information in the lesion image A* to the surrounding fluctuating background B*.
Signal - noise 
In analogy with the analysis of electrical signals in low-current electronics, the terms are introduced for the quantitative description of image properties :
Signal S

is the difference in image intensity
(its "brightness", number of accumulated pulses) between the investigated structure (lesion) and environment. In our case it is given by the difference: S = A*max - B*.
Noise N

represent disturbing statistical fluctuations in the image. For our case, background fluctuations are important, so the noise is given by the square root of the average accumulated number of pulses in the background image: N =
sB = Ö(B*).
   Like the signal quality of the electronics, the quality of the image given by the parameter :
Signal to noise ratio SNR, SNR = S/N = S/
Ö(B*) .
The above statistical condition for the detectability of a lesion can then be expressed as follows: A lesion can be seen in an image only if its signal-to-noise ratio is SNR > 4 .

   If we take into account the effect of resolution and statistical fluctuations, by combining the above relationships we can formulate the basic condition of lesion recognition as follows :
                    C
obj  >  4. e (FWHM / d) 2 / Ö(B*)  .
Only such a lesion will be visible in the image, which will have sufficient contrast C
obj (in the accumulation of radioactivity), geometric size d large enough compared to the resolution of the FWHM camera and the number of accumulated pulses will be large enough so that the relative statistical fluctuations are not too high. The image of the lesion is better the larger, more contrasting the lesion and the higher the density of accumulated pulses in the image. And the smaller the size and contrast of the lesion, the higher we need to accumulate the number (density) of pulses in the image for its successful imaging. For the display of these small and not very contrasting lesions, the best possible resolution of the camera is also crucial in order to avoid an enormous degradation of the contrast of the lesion during the imaging.
Positive and negative lesions 
One of the differences between "cold" (negative) and "hot" (positive) lesions is that the well acumulating hot lesions can have high contrast C
obj even many hundreds of percent, while with cold lesions the contrast can reach a maximum of 100%. Therefore, we can observe well-displayed even the small (but contrastively accumulating) hot lesions, such as inflammatory or tumor foci in classical skeletal scintigraphy or 18FDG PET. Smaller cold lesions are difficult to observe, especially when they are stored deeper (such as inside the liver or lungs).
Deep deposites lesions
 
Phantom measurements in the left part of Fig.4.2.9 (similar to the measurements above in Fig.4.2.7) were performed without a scattering environment (in the air) and near the front of the camera collimator. They simulate an idealized situation of superficial lesions. If the lesion is deposited at greater depths in the tissue, four other adverse factors apply, further reducing imaging contrast and impairing lesion detectability :
v A greater distance from the collimator front leads to poorer resolution (higher FWHM), which reduces the Cimg contrast in the image according to the exponential dependence above.
v
Absorption of g radiation from the lesion as it passes through the tissues (attenuation), reduces the number of useful pulses detected in the lesion image.
v
Radiation from other layers of tissue can be added to radiation g from the lesion. This primarily reduces the contrast of the object Cobj in the respective planar projection and thus the contrast in the image. This effect is largely eliminated in SPECT and PET tomographic imaging (see §4.3 "Tomographic scintigraphy" below).
v Part of the gamma radiation is Compton scattered in the tissue material. Part of this scattered radiation is detected and also reduces the contrast of the lesion image (as shown above in Figure 4.2.8).
   In the right part of Fig.4.2.9 is a phantom display of positive and negative lesions deposited on the surface and at different depths in the tissue (simulated by water with dissolved
99mTc activity). In deep-seated lesions, their image deteriorates sharply, especially in the case of negative ("cold") lesions.
How can image quality and detectability of lesions be improved ? 
The recognizability of small structures (lesions) in scintigraphic imaging is determined in practice mainly by the following factors :
× Geometric size of the lesion;
× Accumulation of radioindicator in the lesion compared to the surrounding tissue ® contrast of the lesion;
× Depth of lesion placement ® attenuation of radiation, interference with radiation from other layers;
× Spatial resolution of the scintigraphic system - contrast in the image;
× Detection efficiency (sensitivity) + acquisition time ® number of accumulated pulses ® statistical fluctuations.
   The size and location of the lesion is determined by the anatomical situation of the patient, the resolution and sensitivity of the camera are basically determined by its construction, but we can partially influence them by a suitable choice of collimator. There are then, in principle, four ways in which we can improve the image quality and the capture of lesions :
l Increase the primary contrast of the lesion
This can be achieved in some cases by choosing a suitable radioindicator, which is more selectively taken up in the diagnosed lesion.

l
Increase the applied activity
of the radio indicator, which will increase the detected number of pulses and reduce the relative statistical fluctuations. However, this encounters the problem of increased radiation exposure of the patient and, at high activities, also for a dead time of detection device.
l Increase the image acquisition time ,
which proportionally increases the number of stored pulses in the pixels of the image and reduces the relative statistical noise. However, too long an acquisition time brings problems with the movement of a patient, which does not last so long motionless under the camera detector. In dynamic scintigraphy, this solution is usually not applicable at all, because the acquisition time of individual images is determined by the time dynamics of the investigated process.
l Perform a suitable computer filtering of the image ,
which can improve its quality and help identify smaller defects
- it is discussed in more detail in the discussion "Filters and filtering of scintigraphic images". It is mainly optimized smoothing of statistical fluctuations ( Low-pass filters - smoothing ) and artificial improvement of resolution - resolution recovery ("Bandpass filters - focusing").
   In general, the lesion in the tissue is displayed more easily, if it is larger, more contrasting and located at a smaller depth below the surface of the body.

Quantification of positive lesions on gammagraphic images - SUV 
One of the most common tasks of radionuclide gammagraphy is to display the accumulation of a suitable radioindicator in lesions (especially tumor) - not only to recognize the lesion in the image, but also to determine the quantitative intensity of radioindicator accumulation in the displayed tissue. A simple relative criterion of the significance of the displayed lesion is the above discussed contrast of the image C
img = (A*max - B*)/B* between the activity (accumulated number of pulses) in the A* lesion image and the surrounding B* background. To assess the severity of tumors in different patients, as well as in monitoring the time course of tumor size and metabolic activity in a given patient (most often monitoring the biological response of tumor tissue to therapy), the degree of accumulation of the relevant radiopharmaceutical in images from various independent scintigraphic studies should be evaluated and compared. For the absolute (semi)quantitative expression of the selective uptake of the radioindicator in the tumor, in comparison with the average distribution in the rest of the body, the so-called standardized accumulation value SUV (Standardized Uptake Value) is often used. It expresses the ratio of the local accumulated concentration of the radioindicator in the lesion to the average concentration in the whole body (ie to the applied activity normalized to the patient's weight) :
                     SUV  =  C / (A
inj / M)  .
Here, C [kBq/cm
3] is the tissue concentration of radioactivity (volume activity) in the lesion, Ainj [MBq] is the applied activity, M [kg] is the mass (weight) of the patient. The values of volume activity of lesion C and applied activity of Ainj must be corrected at the same time (especially when using short-lived radionuclides such as 99m-Tc or 18-F). Concentration C radioactivity in the lesion is determined from the gamma image using the appropriate conversion and correction factors :
                     C  = 
h -1 . (A * -B *). RC -1 .V tum -1  ,
where
h [imp. s-1 MBq-1] is the detection efficiency (sensitivity) of the camera, RC is the so-called recovery coefficient of correction for the "partial volume effect" (mentioned above in the section "Adverse effects of scintigraphy", passage "Partial volume effect"), Vtum [cm3] is the volume of the lesion.
   Therefore, if we measure the value SUV = 1 at some point in the image, the volume activity is the same as the average activity in the whole body - it means that the radio indicator is not captured here. The higher the value of SUV > 1 we get, the more selectively the given radioindicator accumulates in the given place, the higher the metabolic activity of the respective tissue.
   Either the SUV
max calculated from the A*max value of the most intense pixel in the lesion image is used, or the SUVmean (SUV50%) determined from the average value in pixels within the area of interest (ROI) of the lesion, sometimes the SUV70% etc. If there is otherwise an approximately homogeneous distribution of the radioindicator outside the examined lesion, the SUVmax is approximately equal to the contrast value Cimg and other SUV50 or SUV70 values can be determined simply as the ratio of the number of impulses in the tumor (or its defined part - ROI) and in the tissue background ("tumor to background ratio"). However, it is desirable to make a correction to the partial volume effect using the RC recovery coefficients (as mentioned above).
   SUV analysis is performed mainly on PET images of
18FDGs and other radiopharmaceuticals with tumor accumulation - see also §3.6, section "Diagnosis of cancer". For medium accumulating tumor lesion the SUV value is in the range of about 2 ¸ 5, for the well and selectively accumulating tumors can then be SUV> 10.
  
Note : Quantification of SUVs with planar and SPECT scintigraphy performed only quite rarely. SUV is domain a primarily tumor scintigraphy, PET (see below §4.3, section "Positron emission tomography PET"), which is mainly used to quantitate the accumulation of 18F-FDG. Here we discuss the issue of SUV general terms in connection with common properties of scintigraphic images and the information contained therein.
Disadvantages and pitfalls of SUV quantification 
The determination of SUV can be a useful tool for assessing the severity (metabolic activity, possible aggression) of tumors and the effectiveness of the biological response to their therapy. However, it is necessary to keep in mind even some pitfals, consisting in the dependence of the obtained SUV values on a number of circunstances and parameters :

¨
Particular, it is the exact actual value of activity in relation to the calibration of the meter applied activity and sensitivity (detection efficiency) gamma camera. It also depends on the time between application and examination, which by radioactive decay and pharmacokinetics significantly affects the amount of radioindicator accumulated in individual tissues, including the examined lesions. Correction to this time can be difficult because different types of tissues and tumors accumulate the radioindicator at different rates. The only way to minimize this time factor is to keep the same time interval between application and scintigraphy. It is also necessary to subtract the activity values remaining in the syringe or tubing after application.
¨ Hydration and levels of metabolic substances (eg sugars) in the patient's blood, functional state of the kidneys, liver and other organs.
¨ It is also a dependence on the weight and body constitution of the patient. In patients with the higher the fat content, which accumulates very little in the radiopharmaceuticals used, overestimates the measured SUV values. This can be a problem when comparing different patients with each other, or if a given patient changes weight between exams. Correction of SUV to patient weight can be performed approximately by normalization to standard reference values of patient weight 70kg and body surface area S = 1.75m2, using empirical relationship between weight M , height H, body surface area S and adipose tissue fraction: SUVM-corr = C/(Ainj) .43.8.M0.425.H0.725 .0.0072. Due to this correction, the measured value of SUVM-corr in the shown lesions is lower in more massive patients than the uncorrected value of SUV, on the contrary it is higher in more subtle patients.
¨ Marking of areas of interest (ROI) of examined lesions on the scintigraphic image is individually dependent and is not very reproducible. SUV values (especially SUVmean) are very sensitive to small differences in the size and position of marked ROIs.
¨ Absorption of gamma radiation in the tissue, causing attenuation of the signal from deeper lesions (see section "Adverse effects with scintigraphy and their correction"). The correction for attenuation is not always accurate and reliable.
¨ Computer image editing - various kinds of filters, methods and algorithms reconstruction by tomographic images can significantly (and non-linear) influence the accumulated number of pulses in the evaluation of lesions and tissue background. This leads to large arteficial differences in measured values SUV.
¨ The effect of partial volume ( partial volume effect - as described above in the passage " the volume and activity bias ") causes distortion displayed lesions in terms of activity and size. To correct for this effect, RC coefficients are used, which are difficult to determine (obtained by phantom measurements) and their values depend on the imaging properties of a particular camera. To use them, it is also necessary to know the diameter of the displayed lesion.
   Due to these difficulties in determining a specific exact SUV value, this parameter is valid only in comparative studies of larger patient populations, where individual deviations and inaccuracies are randomized. When comparing changes in scintigraphic images in a particular patient, the SUV value
(which in practice cannot be determined with an accuracy of better than 30%) needs to be "taken with a grain of salt"..!..
  
Author's skeptical note on SUV :
The importance of "accurate" absolute quantification of SUVs using all sophisticated correction methods is sometimes overestimated. To gain my own experience, I would like to recommend the following experiment to colleagues : Try to compare the SUV values determined by the above complex procedure, with the values obtained from a simple ratio of the number of pulses from the ROI in the lesion image and the number of pulses in the ROI of a suitable reference healthy tissue. The relative results will be very similar, at least in terms of assessing the severity of the metabolic activity of the tumor and the biological response to therapy..?.. - I welcome your experiences ...
Relative SUV 
These pitfalls of accurate SUV determination show that identical conditions cannot be maintained in practice during repeated scintigraphy of the patient before and after therapy. Therefore
(in connection with the above note) the relative SUVrel is introduced as the ratio SUVrel = SUVtumor / SUVreference tissue , where all problematic values of applied activity, detection efficiency, partial volume effect, patient weight, application time are truncated. We basically get the value of the tumor / background ratio expressing the relative rate of uptake of the radiolabel in the analyzed lesion compared to the tissue background. The SUVrel can be obtained from the scintigraphic image very simply by comparing the number of impulses from the lesion ROI and the ROI of a representative tissue background (eg liver or aorta ROI is used as reference tissue; identical ROI must be observed when repeatedly evaluating the same patient) .

Technical failures of scintillation cameras
With such a complex electronic device as a scintillation camera, there are many possibilities for mild and more serious technical failures. We will mention here only some disorders specific to gamma cameras. In terms of their location, we can divide them into two groups :
1. Disorders of electrical power supplies and mechanical movements of the camera 
Electrical power supplies for cameras are often burdened with long-term power, they become hot, cooling fans "get stuck", ....
   Current gamma cameras in their electro-mechanical part contain a number of sensors, regulation and control circuits, which is certainly correct in terms of successful and safe operation of the device. Sometimes, however, it is too "recombined", so that even an insignificant deviation can lead to blockage of mechanical movements and thus the practical unusability of the camera, with the need for service intervention.
2. Disorders of imaging properties in the field of view of the camera 
Practically all these disorders can be clearly seen in the image of homogeneous distribution of gamma radiation
(whether it is a homogeneoussource, or irradiation of a crystal without a collimator with a point source from a sufficient distance - see "Phantoms and phantom measurements" , section "Testing and calibration of camera image homogenity") :
         
When properly functioning, the image of the homogeneous distribution of radioactivity should also be homogeneous (Figure
a), with the only permissible deviations resulting from statistical fluctuations of the accumulated number of pulses due to quantum-stochastic processes during the emission of gamma photons.
   Local circular outage
(mostly sharply demarcated, often with a visible hem) in the field of view is the result of interrupted detection by scintillation from a specific location of the camera crystal. The cause can be either a failure of the respective photomultiplier, or its preamplifier or some other circuit through which the detected pulses pass. In Figure b is a failure of one peripheral photomultiplier. Repairing a preamplifier is not a bigger problem. However, replacing a defective photomultiplier is technically very difficult. After the electrical disconnection, the photomultiplier must be carefully "peeled off" from the silicon grease (ensuring optical contact with the scintillation crystal light guide), thoroughly clean the area and stick with silicone a specially selected photomultiplier with the same properties as other photomultipliers. This work will take an experienced electronics woker all day, including subsequent adjustments and the resulting calibrations of the camera detector.
   A series of minor inhomogeneities in Figure
c, corresponding to the positions of the photomultipliers, may not indicate a malfunction, but are usually caused by misalignment - "detuning""- positions of photopeak from individual photomultipliers. After proper adjustment - "tuning" - and the creation of a new homogeneity correction matrix, we usually obtain a homogeneous field of view.
   The most serious accident of the scintillation camera is cracked crystal. In the image of the field of view, it appears as a distinctive irregular (zigzak or branched) line of pulse outage, with a positive rim (Fig.
d) This fatal failure can occur in basically two ways :
× By mechanical pressure or impact on a very brittle crystal. It is enough for a screwdriver, phantom holder or other object heavier than a few grams to fall on the crystal without a collimator. When replacing collimators, the crystal may break when there is a foreign object on the mounted collimator, for example a pencil ..!..
× Thermal stress when the temperature of the crystal changes unevenly or rapidly. Larger temperature gradients due to thermal expansion can cause considerable mechanical stresses in the crystal, which can result in cracking. The crystal is particularly temperature sensitive when the collimator is removed. In this situation, it is not even recommended to open windows in the room or turn on the air conditioning.
   A cracked crystal is an irreparable defect in a scintillation camera. The entire detector must be replaced 
(crystal + photomultipliers + preamplifiers) for a new detector, assembled in the factory. This is a costly affair, over $ 100,000 !

New and alternative physical and technical principles of gamma cameras
Practically the only type of scintillation cameras used so far in nuclear medicine are Anger-type cameras described above
(of course with the exception of PET cameras described below in §4.3 on tomographic scintigraphy). Despite the clear success of the use of these cameras in nuclear medicine, two basic disadvantages of this solution were also known from the very beginning. The first is the need to use a lead collimator, through which only the radiation g passes in a precisely defined direction, but the vast majority of incident photons are captured in the partitions between the holes Þ low detection efficiency (sensitivity) of camera. The second disadvantage stems from the limited accuracy with which a system of photomultipliers and electronic circuits is able to locate the position of a scintillation flash in a large-area scintillation crystal Þ imperfect spatial resolution.
   Therefore, since the 1970s, alternative physical-technical solutions of scintillation cameras have been designed and experimentally tested, eliminating the first or second disadvantage, or both at the same time. These alternative solutions have not yet gone beyond laboratory experiments, but with the development of technologies in the field of microelectronics and new materials, there is a real hope in the near future to bring some of these constructions into a practically usable form, or even to replace existing scintillation cameras in the more distant future...

Wired cameras
Wireframe cameras are based on the simple principle of a position-sensitive multi-wire ionization chamber, which was developed for monitoring and displaying traces of particles formed during interactions in accelerators (see §2.3, section "Drift ionization chambers"). The detector itself is made up of a large number (even several hundred) of thin wires - electrodes stretched in a gas charge in two layers in a mutually perpendicular direction - determined by the X, Y coordinates.
When a photon radiation g enters, the ionization occurs at the appropriate site. The electron cloud drifts from this point to the nearest electrodes, where an electrical signal is generated. The intersections of the electrodes thus received signal the location of the interaction of the detected photon. The ionization cloud of electrons can reach several nearby electrodes; the evaluation electronics then determine the coordinates using the weighted averages of the signals from the various electrodes. The point of impact and interaction of the photon can be determined with an accuracy of about 0.1 mm. Cameras of this type are especially suitable for imaging with low-energy radiation g .

Semiconductor multidetector gamma cameras
One of the basic factors limiting the internal resolution of an Anger scintillation camera is the uncertainty with which a system of photomultipliers and subsequent electronic circuits is able to locate the position of a scintillation flash in a large-area scintillation crystal. Therefore, the internal resolution of the Anger camera cannot be reduced below approx. 3 mm in practice.
   The concept of a multidetector camera is that instead of one large-area scintillation detector equipped with a number of photomultipliers, many separate miniature detectors - pixel semiconductor detectors
(see §2.5 "Semiconductor detectors") are used, placed in a matrix next to each other. Gamma radiation is transferred directly here to electrical signals without the need for scintillators and photomultipliers. The signal from each of the detector is processed independently (in multiplexed mode), whereby the positional coordinates (x, y) are determined simply by the position (i, j) of the mini-detector in the detector array, and lead directly into the pixel array in the computer (the pixel to pixel ) - fig.4.2.10 :


Fig.4.2.10. Principle of multidetector semiconductor camera.
Left, center: The crystal of a multidetector camera consists of a large number of regularly arranged miniature semiconductor pixel detectors. Right: Special arc configuration of semiconductor CZT detectors and multi-pinhole collimators for SPECT myocardium.

Photon detection is performed in individual pixels independently, so the internal spatial resolution is given by the size (pitch) of the detector pixels (unlike the Anger camera, where the coordinates of scintillation are determined triangulation according to the response of different photomultipliers). If a sufficiently dense grid of pixel detectors is created, we can achieve a very good internal spatial resolution (even below 1mm); the total resolution then depends on the collimator used. Optimized collimators for multipixel semiconductor cameras should have square apertures the size of pixel detectors (minus the thickness of the baffles), which would geometrically overlap with the detection pixels with their apertures everywhere in the field of view.
  So far, this type of camera has been produced only with a small field of view of about 5
x 5 cm, for a unique use for scintigraphy of small objects (small laboratory animals), now it is beginning to be produced in the standard size of classic cameras. This category also includes electronic imaging detectors for X-rays, so-called flat-panels (described in §3.2, section "Electronic imaging of X-rays", flat panels with "direct conversion", which probably belongs to the future...). Gradually, planar and SPECT cameras of standard dimensions with semiconductor detectors are also being used.
   For this semiconductor gammagraphy
(planar and SPECT "scintigraphy"), semiconductor CZT (Cadmium-Zinc-Tellur) detectors have proven. Cadmium and zinc CZT telluride is a semiconductor detector operating at room temperature, which converts gamma and X-rays directly into electrical impulses with high efficiency (physical aspects see §2.5 "Semiconductor detectors", passage "Cadmium-Zinc-Teluride (CZT) detectors" ). A comparison of the average basic parameters of a standard Anger camera (with NaI(Tl) scintillation crystal and photomultipliers) and a semiconductor camera with CZT detectors (2.5 mm in size) is in the following table :

Camera type Internal spatial resolution Detection efficiency
(for 99mTc)
Energy resolution Max. pulse frequency
Anger camera with NaI (Tl) 4 mm 60 cps / MBq 10% 3 . 10 5 cps
Semiconductor
CZT camera
2.5 mm 85 cps / MBq 6% 6 . 10 5 cps

Thus, compared to conventional Anger cameras, semiconductor CZT cameras have better spatial resolution and energy resolution, slightly higher detection efficiency (sensitivity) and shorter dead time of detection.
   The use of CZT detectors for positron emission tomography of PET is also promising, instead of BGO/LSO scintiblocks with photomultipliers (see below "Positron emission tography of PET", Fig .4.3.5). In addition to better detection efficiency and spatial resolution, a somewhat shorter coincidence time can be achieved (for better TOF). So far, it is being tested experimentally on smaller PET models. The advantage of semiconductor detectors is also theirs independence from the magnetic field, which allows use in hybrid PET/MR systems.
   In nuclear cardiology, stationary semiconductor CZT (Cadmium-Zinc-Tellurid) cameras with a special "cardiofocal" detector arrangement are beginning to be used for SPECT of the myocardium ( Fig.4.4.10 on the right). The detectors are placed in the camera gantry along an arc covering an angle of approx. 90-180°. The detectors are equipped with "multi-pinhole" collimators directed cardiofocally into the center of the gantry. Unlike the classic rotary SPECT
(described below "Tomographic scintigraphy SPECT"), data storage takes place stationary, detectors and collimators are in a fixed position relative to the patient's body, all SPECT projections are obtained simultaneously. This achieves higher detection efficiency and faster processing. However, it is a single-purpose device for SPECT myocardium in cardiology.
   Advanced universal stationary semiconductor SPECT cameras are being developed - see below "
SPECT Stationary Multidetector Cameras".

Compton cameras
In the paragraph on adverse effects of scintigraphy, we classified Compton scattering of
g- rays in tissue as an adverse efect that worsens the quality of scintigraphic images. However, with the appropriate mechanical configuration and electronic interconnection of two or more detectors, Compton scattering g in the detection system itself can, in principle, be used for "electronic collimation" and imaging of the field of radiation g without the use of mechanical collimators (using the Compton scattering for gamma imaging, suggested Everett, Fleming, Todd and Nightengale in 1977). The principle of operation of such a so-called Compton camera is schematically shown in the following figure 4.2.11 :


Fig.4.2.11. Schematic representation of the principle of electronic collimation using energetic-angular reconstructions of the paths of primary (
g ) and Compton scattered ( g' ) gamma-ray photons.

The camera itself consists of two (or several) consecutive detectors providing positional and energetic information about the detected quantum g :
  In the first thin detector 1 (replacing the classical lead collimator) occurs a Compton scattering of photons of incoming radiation
g (by different angles J), which then continue their movement to the second more massive detector 2, where they are fully absorbed.
In the coincidence mode, the positional coordinates of the impact of the primary photon
g (x1, y1) and the energy E1, transmitted to the electron at Compton scattering in the first detector are detected, as well as the positional coordinates of the impact (x2, y2) and the energy E2 of the Compton scattered photon g' absorbed in the second detector. Based on the geometric comparison of the positions (x1, y1) of the primary and (x2, y2) scattered gamma photons, the angle J of the compton scattering is determined. This angle J is then related to the energy E1 of the Compton scattering and the energy E2 of the scattered radiation g', which allows (according to the relation for the angular-energy distribution of Compton scattered radiation Eg ' = Eg / [1 + (Eg / moe c2). (1 - cos J)], given in §1.3) to kinematically reconstruct the path of the photon to determine the incidence angle j at which the primary photon g flew to the first camera detector from its source. Photopeak measurement Eg = E1 + E2 then it makes it possible to eliminate those unwanted photons which were scattered by Compton before coming to the first detector, similarly to Anger's camera.
   This creates an incident cone with a vertex at (x
1, y1) and an apex angle J, on the mantle of which lie the possible trajectories of the incoming photon. The set of these incident cone shells from individual detected photons can then be used for computer reconstruction of the resulting scintigraphic image of radioactivity distribution in the scanned object: in the matrix of the reconstructed image are summation occupied the "pixels" corresponding to the intersection of individual conic sections (ellipses, circles), arising from the projection of incident cones into the plane (Fig.4.2.11 on the right is an example of the reconstruction of the image of a point source, arising as an intersection of elliptical projections of incident cones of photons emanating from this source).
   In the scattering detector 1, a multidetector system of semiconductor detectors Si, CdTe or GaAs with a thickness of about 5 mm is used, a high effective cross section for Compton scattering is required here. The absorption detector 2 can be an Anger crystal system NaI(T1) or BGO or LSO with photomultipliers and electronics evaluating the position of the flashes. However, even in this second detector, it is advantageous to replace the Anger camera with a semiconductor multicrystalline detector. In addition to spatial and energetic resolution, for the good operation of the Compton camera, high demands are also placed on the temporal resolution of the coincidence (similar to PET detectors - see §4.3).
   Compared to mechanical collimators, electronic collimation can lead to a significant improvement in detection efficiency (sensitivity), as
g photons are used from a much larger spatial angle (electronic collimation, but of a different kind, is of great importance in positron emission tomography, see PET below).
   Apart from laboratory experiments, Compton's cameras have not yet been implemented, they will probably remain only a physical-technical interest....

High-energy gamma cameras
The need to imaging the distribution of high-energy
g radiation arises mainly in two areas :
× 1. Gammagraphic imaging of the distribution of radioactive substances emitting hard gamma radiation (their distribution in samples, tissues and organs), or depicting the distribution of atomic nuclei excited by external radiation that emit high-energy radiation g during deexcitation (such as the NSECT method - see "Neutron- stimulated emission computed tomography" below). However, this is a relatively marginal issue ...
× 2. Gamma-telescopic imaging sources of gamma radiation in space - supernovae, neutron stars, accretion disks around black holes (see eg "Astrophysical significance of black holes" in the book Gravity, black holes and space-time physics ) and other turbulent astrophysical processes; on g radiation from space, see also §1.6 "Cosmic radiation", section "Cosmic X and gamma radiation".
   Imaging with high-energy gamma rays - hundreds of keV to tens of MeV - is much more difficult than with soft g -radiation (60-500keV). For such energies, the collimators have poor spatial resolution and luminosity due to the significant cross-radiation trough the septa between the holes, and the scintillation crystals of standard gamma cameras used are too thin to achieve reasonable detection efficiency. A suitable solution here is the above-mentioned principle of the Compton camera, in a modification optimized for high energies. A simpler type of Compton telescope, used on space stations to detect g- rays from space sources, consists of a larger ionization chamber (drift wire or projection) in which are measured the energy of the scattered g- rays and reflected electrons, as well as the direction of scattered radiation or reflected electrons.


Fig.4.2.12. Some principles of gamma cameras for high energy.
Left: Combined Compton-Anger high energy gamma camera. Right: 3-Compton gamma-telescope with many detection layers.

Fig.4.2.12 on the left schematically shows the principle of operation of a combined Compton-Anger gamma camera for high energies. Detetion sensitive camera volume consists of ionization drift-time projection chamber (TPC - Time Projection Chamber) with a gas filling (the ionization detector, see §2.3 "Ionization Chamber"). When a high-energy photon g flies into this working space, a Compton scattering in the gas filling mainly occurs, for higher energies also the formation of electron-positron pairs, followed by annihilation of a positron with an electron to emit a opposite - pair of gamma photons with energies of 511keV. The path of reflected or paired electrons is sensed based on the ionization electrons that these high-energy particles generate along their paths. They are detected by a matrix of several hundred miniature pixel ionization chambers, working in proportional or Geiger (avalanche) mode. Or semiconductor detectors can be used. This cell matrix forms a 2-D position sensitive electron detector. The ionization electrons from the individual paths of the fast charged particle drift into different chambers (perpendicular projection of the path into the nearest chambers) for different lengths of time; by evaluating these geometric and temporal data, the 3-D path of reflected or paired electrons and positrons in the chamber space can be reconstructed. The working chamber is surrounded on all sides by scintillation crystals with photomultipliers (Anger camera), scanning Compton scattered and annihilation photons, with scintillation positioning and radiation energy. By a complex coincidence evaluation of pulses from the matrix of pixel detection chambers and from the photomultipliers of the Anger camera, it is then possible to geometrically reconstruct the direction (angle) from which the detected primary high-energy photon g arrived - to realize gamma-ray imaging .
   Fig.4.2.12 on the right shows the principle of a gamma-telescope based on repeated Compton scattering in layers of position-sensitive semiconductor detectors. The system consists of several layers of flat position-sensitive (2-D) detectors, stacked at equidistant distances. After the entry of the primary
g -photon with energy Eg1 is scattered by Compton in one of the detectors, which is accompanied by a position pulse and an amplitude pulse carrying information about the energy loss DE1 that the photon left in the detector during scattering. The scattered photon continues to fly at an angle J1 with energy Eg2 = Eg1 - DE1, after which it can be further scattered by Compton in another detection layer, providing the appropriate position pulse and energy pulse DE2. The photon scatters by an angle J2 and continues with the energy Eg3 = Eg2 - DE2. Thus, repeated multiple scattering can occur until the photon leaves the detection space. Coincidence evaluation of position coordinates in individual detection layers determines scattering angles J1 , J2 , J3 , ...., evaluation of pulse amplitudes determines energy losses DE1, DE2, DE3, ... These data substitute into Compton 's equations
        E
g2 = Eg1 /[1 + (Eg1 /moec2).(1 - cos J1)]   ;   Eg1 = DE1 + Eg2 = DE1 + {DE2 + [DE22+ 4moec2.DE2/(1-cosJ2)]1/2/2} , .... ,
which allows kinematic and geometrical reconstruction of the photon path - will provide the required value of the angle
J of the incident cone, under which the primary g- photon flew into the detection system. And further reconstruction by intersecting a set of projections of incident cones of all registered photons, the resulting g- telescopic image of the source from which the photons were emitted is obtained. The advantage of this arrangement is that to reconstruct the angle of incident g-photon does not need its complete absorption in a heavy "calorimetric" detector to determine the total energy. The energy of the incident photon is determined by measuring the position of the first three interactions and the energy delivered in the first two interactions.
   Thus, it is sufficient to obtain at least a 3-fold Compton scattering, the analysis of which can be used to reconstruct the incidence angle
J - hence this system is sometimes referred to as the 3-Compton telescope. The analysis of possible further scatterings refines the reconstruction. In the individual layers of flat position-sensitive detectors it is possible to use either ionization wire chambers, or better germanium or silicon semiconductor drift detectors, which have good energy and image resolution.
   To imaging gamma radiation very high energies, hundreds of MeV to hundreds of GeV, special particle detectors of electron-positron pairs are used in an arrangement similar to Fig.4.2.12 on the right. The
g- rays first fall on a plate of heavy material (tungsten), where they are converted into electron-positron pairs, flying almost in the direction of the original photon g. Their paths are then monitored by layers of position-sensitive 2-D silicon detectors (trackers), which determines the direction from which gamma radiation came. Finally, they transfer their energy to a calorimetric detector located below the last detection layer, which detects the energy of the g- quantum.


4.3. Tomographic scintigraphy
Every living organism is a three-dimensional object and the distribution of a radioindicator has the same character. A planar scintigraphic image, which is a two-dimensional projection of an object, can therefore capture only part of reality. From the planar scintigraphic image we cannot find out anything about the distribution of the radioindicator in the "deep third dimension", perpendicular to the front of the collimator. Planar scintigraphic images have serious pitfalls in this respect - the possibility of overlapping and superposition of structures stored at different depths. Althout we help here by displaying in several different projections, the risk of a false finding or non-detection of an anomaly in the depths of the organism, covered by another structure, can never be ruled out. The superposition of radiation from different depths of the imaged organism further leads to a reduction in the contrast of the imaging of the lesions, which are overlapped in the planar image by radiation from the tissue background thus formed.
  To overcome these disadvantages of planar scintigraphy and to obtain a complex image of structures at different depths, tomographic scintigraphy *) has been developed to provide a three-dimensional image of the radiolabel distribution. One of the main advantages of tomographic imaging is significantly higher contrast imaging of lesions (up to 10-fold) that do not overlap with tissue background radiation on transverse sections.
*) Greek tomos = section - the tomographic image consists of certain sections, mostly transverse, a larger number of which create a three-dimensional image.
   Some basic principles, especially geometric and reconstructive, have all tomographic methods in common. X-ray transmission tomography CT was described in §3.2 "X-ray diagnostics", part "
Transmission X-ray tmography (CT)", where the development of tomographic methods in general is also mentioned.
Technical development of gammagraphic tomography 
Efforts to achieve in-depth tomography imaging began shortly after the introduction of scintigraphy in the 1960s and 1970s. The foreruner of contemporary gammagraphic tomography SPECT in the 1970s was movement tomography (Fig.4.3.1 left): the examining table with the patient and the collimator of camera with inclined holes (slant holes) using the electromotor synchronously rotated in such a way that, for a layer in a "focal" depth, both movements were compensated and a sharp image was obtained, while in the other layers (above and below the focal plane) the image was motion blurred and thus it was darker and less distinct. Against the background of these blurred and darkened areas, sharper and more clearly displayed structures from the focal plane were better visible. The depth position of the focal plane was set by the radius of rotation of the lounger on the eccentric of the lounger motor. However, the quality, contrast and depth effect of such an image were not great (completely incomparable with SPECT). An image of only one longitudinal layer was obtained at a time, in order to create an image of another focal layer, it was necessary to change the radius of the sliding rotation of the lounger and start a new acquisition. More detailed tomographic imaging in multiple layers was therefore time consuming. This method has long been abandoned.


Fig.4.3.1. Early attempts to implement tomographic scintigraphy.
Left:
Movement tomography with rotating slant-hole collimator and rotating lounger. Middle: Coincidence tomography using
g-g angular correlation. Right: SPECT on a stationary planar gamma camera with patient rotation.

  Interesting experiments were also performed with 2-photon coincidence tomography using g-g angular correlations between the emission of cascade pairs of gamma photons in some radionuclides (§1.2, part "Gamma radiation", passage "Angular correlations of gamma radiation") - Fig.4.3.1 in the middle. Another gamma detector, or in a more advanced version another gamma camera with a slit collimator, in a coincidence connection was attached to the basic imaging gamma camera at a certain place and at a suitable angle q (eg 90°). To create the image, only pulses originating from the simultaneous arrival of a pair of cascading gamma-quanta into both detectors were registered. Thus, only one longitudinal thin layer was displayed, defined by the detection angle of the auxiliary coincidence detector (or a larger number of independent layers when using a coincidence camera with a slit collimator). However, the palette of suitable isotopes exhibiting cascade deexcitation with angular correlation of the gamma-photons emitted is very limited and the detection efficiency of the coincidence system has been very low. This method did not go beyond laboratory experiments; however, in a sense it can be considered the ideological forerunner of positron emission tomography. Namely, only a perfect angular correlation of 180° between a pair of annihilation photons in electron-positron annihilation, it has found wide application in coincidence positron emission tomography, see PET below.
   The first attempts at SPECT gamma tomography were performed in the 1960s and 1980s at a some workplaces with planar Anger cameras
(the first tomographic image was presented by Kuhl and Edwards in 1963). Since the (planar) cameras at the time did not have gantry and could not rotate, the patient turned - Fig.4.3.1 right. In front of a vertically set stationary camera, the patient sat on a swivel chair with a marked angular scale (goniometer). A planar image was taken, the chair and the patient were rotated by a certain angle, another image was accumulated, etc. - approx. 16-64 images for a sequence of angles 0-360°. This was followed by computer reconstruction by back projection into transverse sections. This method, despite its mechanical clumsiness, has in fact already enabled full-fledged SPECT imaging - with the limitation that at that time not yet sufficiently complex software for the reconstruction of transverse sections and their processing had been developed.
   The mouting of gamma cameras on gantry with the rotation of the camera around the patient then became a truly successful and routinely used method of SPECT tomography (as described below). Recently,
stationary SPECT multidetector cameras without rotation have been developed, which is likely to replace the clumsy rotating SPECT.
Author's note: 
In the 1970s , we also performed early attempts at tomographic imaging according to Fig.4.3.1 at our Department of Nuclear Medicine KNsP in Ostrava-Poruba. Our first Pho Gamma HP gamma camera (Nuclear Chicago) with the CLINCOM evaluation device from 1974 was equipped with a rotating Slant Hole collimator and a rotating lounger for motion tomography (Fig.4.3.1 on the left). Experiments with
gg coincidence tomography we performed on a Pho Gamma HP camera with the help of a perpendicularly oriented collimated scintillation probe, connected in coincidence to the flow of scintigraphic pulses (we did not have an additional gamma camera with a slit collimator). We also tested the improvised SPECT with a patient in a swivel chair on a Pho Gamma planar camera, with the then latest computer evaluation device GAMMA-11 and our own developed software.
   Unlike X-ray transmission tomography CT, where the image is created by the passage - transmission - of X radiation through the body, scintigraphic tomography creates the image by detecting radiation emitted from a radioindicator inside the body. Radionuclide emission computed tomography (ECT) is of two types :
1. Single-Photon Emission Computerized Tomography SPECT, using g- radionuclides registers only one emitted gamma radiation photon from each radioactive transformation .
2. Two-photon Positron Emission Tomography of PET, using positron (beta+) radionuclides, resp. the resulting annihilation radiation, where two photons emitted during annihilation are always detected at the same time (coincidentally).
   In both cases, the resulting tomographic images are obtained by computer reconstruction after scanning the pulses from the photon detection. We will discuss both tomographic methods in this order 1. , 2.


Tomographic scintigraphy SPECT
The most common method of tomographic scintigraphy is the so-called single-photon emission computed tomography SPECT (Single Photon Emission Computerized Tomography). Its principle is shown in Fig.4.3.2 :


Fig.4.3.2. The principle of capturing scintigraphic images of the examined object W (here the brain) at different angles by a rotating SPECT camera (left) and their computer reconstruction into the resulting image W* of a cross section of this object (right).

The basic principle SPECT
The SPECT tomographic camera differs from the conventional planar camera only in that the special stand on which the camera detector is mounted, so-called gantry circular shape
(Gantry = portal, through load -bearing supporting structure), allows the motor-driven rotation of the detector around of the examined object *) - the photograph of the SPECT camera is above in §4.2 in Fig.4.2.4 at the bottom right. To speed up the acquisition, two detection systems ("double-headed" SPECT camera) are most often installed on the common gantry (there were also cameras with 3 or 4 detectors, but it didn't prove very well in practice...).
*) Occurs rarely also the technical construction of a SPECT camera without a gantry. The camera detectors are mounted on special arms equipped with servomotors, allowing the detectors to move in space in different directions and at different angles - with all "degrees of freedom". By suitable electronic control of the servomotors, it is then possible to achieve a circular movement of the detectors around the examined object (around the bed with the patient) during SPECT.
   The new stationary SPECT multi-detector cameras have a completely different principle .
Acquisition of SPECT 
Own tomographic scintigraphy SPECT then takes place in such a way, that the camera gradually orbits *) around the examined object and at a number of different angles captures (planar) scintigraphic images of the examined object - the number of these projections is usually 32, 64 to 128 images at angles 0°-360° - Fig.4.3.2 left. [ Note: In some cases, a smaller range of angles is used - some projections, the quality of which would be degraded by increased absorption (attenuation) of g- radiation and would not contribute to the resulting reconstructed images (they could rather cause deterioration), are not captured. Such is the situation with myocardial SPECT , which is sensed by a range of angles from 90 to 270°, while the angles between 0-90° and 270-360°, corresponding to the rear and right side projections, the images due to a significant attenuation of radiation g are not taken.]
*) The
SPECT stationary multi-detector cameras without rotation are mentioned below.
   Orbiting detector camera around the object to be examined is usually a stepper ( step-by-step ) - the camera rotates a certain angle and stops, the corresponding projection is acquired for a preset time, then it rotates again by a given angular step and the next image is acquired. Continuous rotation of the detector with continuous acquisition is rarely used. From the geometric point of view, may detectors orbit circular - which is used more often, or elliptical (non-circular), with event. using the "auto-countouring" system (mentioned above in the section "Design of scintillation cameras"), in order to better "copy" the body surface and keep the shortest possible distance of the camera collimator face from the displayed structures (to achieve the best possible spatial resolution).
Collimators for SPECT 
Orbiting cameras during the acquisition of SPECT are usually equipped with standard collimators with parallel holes, the same as used in planar scintigraphy: for
99mTc it is mostly HR collimator, for 131I HE collimator, for 123I MediumEnergy collimator. Sometimes they are used special collimators with a different hole geometry: the FanBeam collimator for SPECT of the brain, or a convergent collimator for SPECT of the myocardium. All of these types of collimators have been described in more detail above in the "Scintigraphic Collimators". For stationary multidetector SPECT cameras are used Pinhole type collimators (or special mechanically movable collimators which, however, it is only a temporary solution ...).
Reconstruction of SPECT images
 
From this series of planar scintigraphic images - projections - taken at different angles
(these are planar projections of the distribution of the radioindicator to different angles) are then computer reconstructs the image of the distribution of radioactivity in the imaginary cross section, guided by the examined object in a plane perpendicular to the axis of rotation of the camera - Fig.4.3.2 on the right. SPECT computer reconstruction methods are described below in the section "SPECT computer reconstruction".
   Such a reconstruction can be performed for each row of the image matrix of angularly scanned images, so that a whole series of "stacked side by side" cross-sectional images is created in the computer's memory - a kind of three-dimensional "cylinder"
(for cameras with circular field of view) or "cube" (resp. square - for cameras with a quadrangular field of view), representing a three-dimensional image distribution of a radio indicator in the examined object. The cells of this three-dimensional image already have a volumetric character and are called "voxels" (volume-pixels).
   With this three-dimensional image in computer memory, we can use computer graphics methods to guide and display sections in any direction on the monitor - not only primary transverse sections, but also longitudinal and oblique sections, we can make various geometric reorientations and other adjustments to show the desired structure as clearly as possible. Using computer graphics methods, three-dimensional 3D images can be created using suitable shading and perspective angular display with a number of computer effects, which are artificial and may not directly reflect reality, but are very illustrative and effective, also for didactic purposes.


Example of 3D-imaging in myocardial scintigraphy.

SPECT stationary multidetector cameras
The basic disadvantages of standard rotational scanning of scintigraphic projections at SPECT are clumsiness, slowness and low detection efficiency. At each angle, only a small portion of the gamma photons
(going in the direction of the current position of the detectors) are registered, the other photons emitted from the patient's body are lost. The detectors slowly move to more and more angles; with a very long acquisition time, the number of accumulated pulses in the images is relatively low. We will obtain the required SPECT images only ext post, after end the measurement and computer reconstruction. For the rotation method it is not possible to perform dynamic SPECT scintigraphy (except for myocardial perfusion, where this is allowed by ECG-gating), or to operatively modify the acquisition procedure.


Fig.4.3.3. The stationary SPECT camera detects projections from all angles simultaneously using a large number of circular multipix detectors. There is no any rotation. The resulting data can be continuously reconstructed into cross-sectional images.

To eliminate this disadvantage of rotary SPECT, it is offered to use a larger number of smaller imaging stationary detectors - gamma cameras, placed in a ring around the examined object, without rotation - Fig.4.3.3 left. All projections are then taken simultaneously from all angles. The camera is compact, it does not contain any moving parts, the use of photons for imaging is much more efficient - even with lower applied activity, SPECT examination takes significantly less time. Interfering mechanical phenomena, such as displacements of the center of rotation, do not apply here. Cameras with multiPinhole collimators, multiDivergent, or multiParalel collimators are being developed.
   Here, SPECT tomographic images can be reconstructed and displayed continuously during the acquisition, Fig. 4.3.3 on the right
(similar to planar scintigraphy and PET ), which also enables dynamic SPECT scintigraphy. This type of camera is currently being used in nuclear cardiology for SPECT myocardial perfusion (§4.9.4, section "Scintigraphy of myocardial perfusion", see also section "Semiconductor multidetector cameras" above). Stationary compact SPECT cameras with multipixel semiconductor detectors (described above "Semiconductor multidetector gamma cameras") undoubtedly belongs to the future, surely sooner or later it will push out cumbersome SPECT cameras with rotating detectors ..!..

Computer reconstruction of SPECT
Sinogram, Linogram 
If we successively draw images taken at different angles
J, the individual points of the object in them will describe circles with different radii (according to the distance from the center of rotation). X and Y - coordinates of these circular orbiting points will describe sinusoidal (or cosine) curves R.sin J with amplitude given by distance R from the center of rotation. Their brightness is modulated by the accumulated number of pulses in individual pixels of angular projections. The set of all these graphically represented coordinates of the curves of circulation for all points in a given section creates an important 2D image called a sinogram. It is created simply by gradually taking selected lines (corresponding to a given transverse section) from the individual projection images and storing them "on top of each other" in a new image - a sinogram. This is done for all projections (angles J). Sinogram has two roles in tomographic scintigraphy :
1. It is used as a data format (sinogram-file), into which the acquisition of primary data from individual projection angles takes place at predefined times per image. Some new gamma cameras (especially PET) can operatively store data even in LIST mode (without predefined time per frame), from which the sinogram can be additionally created, with the possibility of computer editing. Each row of the acquisition matrix - each transverse section - has its own sinogram. Cross-sectional images (using the inverse Radon transform) can then be reconstructed from the sinogram data.
2. The sinogram display allows you to check the correct course of the tomographic examination. Under normal circumstances, the sinogram of each displayed active site must have a smooth uninterrupted course ("wavy line"). Any patient's movement (which may lead to deterioration or artifacts) during SPECT acquisition is clearly seen on the sinogram as a discontinuity in the smooth course. Sinograms are also used to test mechanical disturbances during the rotation of detectors, such as displacements of the center of rotation (see below "Adverse effects of SPECT and their correction", passage "Mechanical instability of the axis of rotation"). Based on sinograms, unwanted movements can be corrected by software.

Examples of sinograms (top) and linograms (bottom) in SPECT scintigraphy. In the middle part there are transverse sections .
a) A point source at a distance R from the center of rotation on the sinogram describes a sinusoidal curve x = R.sin
J. The weak auxiliary source at the center remains motionless on the axis of rotation (x = 0).
b)
The displacement of the source in the radial direction during the acquisition is reflected in the discontinuity on the sinogram.
c)
Sinogram of 5 active lesions in Jasczak phantom.
d)
Example of sinogram and linogram of the SPECT brain for imaging dopamine receptors ("Scintigraphy of receptor systems in the brain")

In tomographic imaging methods (SPECT, PET, or MRI), a so-called linogram is sometimes constructed - an image in which the summed rows from all primary images are "stacked" as columns linearly side by side. In SPECT, a linogram is created by summing all rows in each of the projection images at individual angles and saving the resulting summation row as a column in a new image - the linogram. This is done for all projections. Unlike sinograms, of which there are a large number (for each transverse section), the linogram is one for the entire tomographic examination. Computer methods for reconstruction of cross-sectional images (inverse Radon transformation) have been developed by integration along the lines in the linogram. In some cases, the linogram may also be used to assess the smooth running of the SPECT examination with respect to movement artifacts or transient electronic disturbances in the acquisition process.
Note: Universal tomographic regularities 
The regularities and relations between circular planar projections at angles
J and transverse tomographic sections (sinograms, Radon transformations, reconstruction methods, formulas in Fig.4.34), outlined in this part, apply not only to SPECT scanning by physical camera rotation, but also to stationary SPECT, when PET, CT or NMRI. They are basically universal and are used in various imaging modalities.
SPECT reconstruction methods
 
The amount of data accumulated in individual projections at different angles, implicitly contains information about the spatial depth distribution. In order to be able to explicitly display the depth spatial distribution of the radioindicator in cross sections, it is necessary to perform a computer reconstruction of the accumulated "raw" data. Two methods of computer reconstruction of accumulated planar images from different angles to the desired transversal sections are used :

1. Back-projection method
The analytical back-projection method by computer simulates an inverse process to acquire the SPECT scintigraphy: As if the camera detector emitted rays of radiation - of an intensity modulated by the image (accumulated information in the cell) - from each position (the angle where it was rotating and from which it accumulated the relevant image) and from each of its cells, back towards the object under investigation, where these rays "draw" a cross-sectional image in an imaginary image matrix located at the center of rotation. The information contained in one given pixel of the image stacked from a certain angle is transferred to all the pixels of the emerging cross-sectional matrix located in a line perpendicular to the detector. Different "intense" rays from different angles then "irradiate" and "occupy" individual elements (cells, pixels) with differently large numbers as they pass through the created image matrix, which add up when passing through other rays (from other angles). In places where most rays of higher intensity pass, "hot" places with a high accumulation of impulses are created - they correspond to places with a higher concentration of radioindicator in the examined object.


Fig.4.3.4. The process of acquisition of SPECT and reconstruction of the transverse section by the method of filtered back projection.

Thus, this method uses the back projection of data from individual planar images into the originally empty matrix, always at the angle at which the planar image was created. The resulting matrix - the reconstructed image - is created by direct addition of these projection data. Simple back projection has the disadvantage of a higher disturbingly structured background with the formation of "star-shaped" artifacts (see below). In practice, therefore, filtered back projection FBP (Filtered Back Projectoion) is used, which is a variant of the inverse Radon transformation, in which suitable filtration is included. The relevant mathematical formulas are shown in Fig.4.3.4, where the whole process of acquisition, filtration and reconstruction of SPECT is shown from a mathematical point of view (3rd dimension is omitted) :
   The examined object (patient), whose cross section has the distribution of the radio indicator A(x,y), is captured by the camera in a series of projections at different angles J, thus creating images of projections p(u). These images are then Fourier transformed into the frequency domain and the resulting spectra p(n) are multiplied by a filter composed of a RAMP filter and a user filter (see "Filters and filtering"). The created filtered spectra pF(n) are then converted back to the spatial region by inverse Fourier transform (filtered images of projections pF(u) are formed), after which by the back projection (at the same angles J) the resulting image of cross section f (x, y) is formed.
   The filtered back projection method is the most used because it is relatively fast
(fast Fourier transform algorithms are used, the values of trigonometric functions are calculated in advance for discrete values of angles, so that common arithmetic operations are then used). However, in terms of the relationship between the actual distribution of radioactivity A(x, y) and the reconstructed cross-sectional image A´(x, y), it is not exactly a mutually unambiguous representation - the image is constructed not from local values in pixels, but by superposition of projection rays. These projection beams are artificial and leave traces in the resulting image, that do not correspond to the actual distribution of radioactivity in the object under investigation. This is most pronounced in the vicinity of foci of increased deposition of radioactivity, where converging projection beams form a "star-shaped" artificial structure - the so-called star effect. Although this star artifact is effectively suppressed by a RAMP filter, various "noodles" or "filaments" are always visible in the images reconstructed by back projection (below in the figure on the left). These disadvantages of retrospective projection are largely eliminated by iterative reconstruction method. The RAMP filter, which also acts as a "focusing" (emphasizes details and edges in the image), is used in the reconstruction in combination with a user "smoothing" filter to reduce statistical fluctuations, noise, in the image. By a suitable choice of this filter and its form-factors, it is possible to achieve optimal contrast, detailing and noise reduction (it is discussed in more detail in the work "Filters and filtration").
    

2. Iterative reconstruction method
Iterative method *) of reconstruction seeks, by means of successive steps - approximations - such a cross-section image that would best correspond to the individual scanned projections at different angles
J. It is an algebraic reconstruction technique (ART).
*) The Latin word "iteratio" means "repetition"; these are recurring cycles of successive approximations.
   Iterative reconstruction takes place in the following stages :

  1. The initial (default) estimate of the cross-sectional image is determined - 0. approximation. This initial approximation can be basically arbitrary - even zero or constant values at all points. However, for the rapid convergence of further iterations, it is more advantageous for the initial approximation to at least partially resemble the actual image. Therefore, the image created by back projection (described above) is most often used as the default, which we know is usually a good approximation of the real image.
  2. By comparing the mathematically simulated projections of this image with the actually scanned projections for the individual angles J, the corresponding deviations for the individual cells of the image are determined.
  3. Based on these differences, the appropriate corrections are made at the points of the previous image - the 1st approximation is created.
  4. Points 2 and 3 are repeated cyclically - these iterations gradually create the 2nd, 3rd, ...., n-th approximation, which should better and more accurately describe the actual distribution of the radio indicator in the cross section of the examined object.

The iterations are repeated until a certain (preset) convergence criterion is met, such as the required accuracy or a preset number of iterations.
  It could be expected that as the number of iterations increases, the overall image quality will increase. However, experience shows that this is true for about 4-8 iterations. A higher number of iterations then only increases the statistical fluctuations, i.e. the signal-to-noise ratio worsens.
  Compared to the back-projection method, the iterative method has the basic advantage **) that no star artifacts are formed (RAMP type filters are not used here). Also in areas with low radioactivity (near background), the cross-sectional images are "cleaner" and more contrasting - they do not contain "filaments" as remnants of backscatter beams. Another advantage of iterative reconstruction is the possibility of introducing some corrections directly into the reconstruction algorithm - correction for collimator properties, dependence of resolution on distance from collimator ...
**) However, "no wonders" can be expected from the iterative method of reconstruction when processing SPECT images in routine clinical practice. - the difference from the back-projection method is often not even noticeable, as the image quality is primarily due to insufficient statistics, camera resolution, scattering and other disturbances (mentioned below) with which no reconstruction method "will do nothing"...
  The iterative method of reconstruction is much more demanding on the number of arithmetic operations, so it could be routinely used only with the development of sufficiently fast computers (using coprocessors) with a high memory capacity.
Improved variants of iterative reconstruction 
In order to streamline and speed up iterative reconstruction, some newer variants and modifications of the basic iterative procedure have been developed :
EM (Expectation Maximalization) - finding the best estimate of the image by statistical methods ........
ML (Maximum Likekihood) - estimating maximum likelihood principle ......
MLEM
(
Maximum Likelihood Expectation Maximization) - iteration procedure with a preset number of iterations: before the start of the reconstruction, the number of iterations is preselected for which we assume the optimal image quality.
OSEM
(Ordered-Subset Expectation Maximalization) - the set of all projections is first regularly divided into several smaller groups (subsets) and the iteration step is applied to individual subsets separately. The sub-iteration of each subset serves as an input estimate for the iteration of the next subset. One complete iteration step is an iteration cycle across all subsets. The product of the number of subsets and the number of iterations in each of them determines the effective number of iterations . From a computational point of view, the OSEM method is approximately as many times faster as the number of subsets we use.
SART(Simultaneous Algebraic Reconstruction Technique) - works simultaneously on multiple sections of a 3-D image ............
OSSART - combination of OSEM and SART methods ............
... ..... add .........
Hybrid reconstruction of SPECT-CT ? 
Some new SPECT / CT hybrid systems attempt to improve the quality of SPECT images through special iterative reconstruction, integrating SPECT and CT data during reconstruction using local ("zone") CT density maps of soft tissues, lung or adipose tissue, and bone tissue. These zone CT density maps define tissue boundaries and modulate their coefficients ("remodel") the primary scintigraphic data of the radiotracer distribution. This achieves a sharper boundary of bone tissues and lesions - provided that these tissues take up the radiopharmaceutical (eg 99m-Tc phosphonates). This modulation may also more significantly show the differentiation between cortical and spongy bone in the vertebrae and flat bones, or between the cortex and cavity in the long bones. Simply put, modulation by CT coefficients gives the SPECT images of the radio indicator distribution a higher contrast .
  A slight improvement in the quality of the images is visible, but the model dependence is debatable here (confrontation with classical reconstruction is desiderable!).

Advantages of SPECT
Compared to planar scintigraphy, SPECT tomographic scintigraphy has three advantages :
¨ More precise determination of the anatomical position of structures and their shape in a three-dimensional image, when viewed from different angles.
¨ Better separate display of lesions lying behind each other at different depths.
¨ By suppressing the superposition of radiation from overlapping layers, a significantly better image contrast is achieved, which enables more sensitive recognition of small lesions even at greater depths.

Adverse influences when SPECT and their correction
As mentioned above, the main advantages of tomographic scintigraphy are the provision of a clear complete image "from all sides" (3-dimensional image) and a significantly higher contrast of the imaging of the lesions against the tissue background. However, we will also mention some disadvantages and pitfalls of SPECT scintigraphy.
  As with planar scintigraphy, SPECT scintigraphy has some adverse and disruptive effects that may degrade imaging quality. Here are six basic adverse effects, the first three of which are also known from planar scintigraphy
(however, with the SPECT method they manifest themselves more markedly and in a slightly different way), the last three are specific to SPECT (rotary method).

Note: About the pitfalls and possible errors of correction methods, bassicaly here the same applies as in general scintigraphy - described above in section "Errors and pitfalls of correction methods, correction artifacts".

Use of SPECT scintigraphy
SPECT tomographic scintigraphy represents a significant addition
(relative to the planar scintigraphy) and improvement to the geometric-anatomical information on the distribution of the radioindicator in tissues and organs. It is mainly used in scintigraphy of myocardial perfusion (§4.9.4 "Scintigraphy of myocardial perfusion") and brain (§4.9.8 "Perfusion scintigraphy of the brain"), as well as receptor scintigraphy of the brain ("Scintigraphy of brain receptor systems"). Also in other scintigraphic methods, such as skeletal scintigraphy or tumor localization diagnostics, tomographic SPECT imaging is beneficial.
  
Another important possibility, the specification of anatomical localization, is the fusion of SPECT + CT images in two-modal combinations of SPECT/CT (see below the section "Fusion of PET and SPECT images with CT and NMRI images"). Scintigraphic images provide important information about the functional status of tissues and organs, but are usually unable to provide sufficient anatomical information about the exact location of pathological abnormalities (lesions) imaged scintigraphically. Radioactivity does not enter the surrounding anatomical structures (eg skeletal), which do not capture the radioindicator and are not visible in the scintigraphic image. It is therefore optimal to perform a better and clearer comparison of the character, size and location of the displayed structures while simultaneous imaging of SPECT + CT or PET + CT images, where X-rays of CT provide precise anatomical localization of the examined structures.


Positron emission tomography PET
Positron Emission Tomography (PET) is a method of scintigraphic imaging of the distribution of positron (b+) radionuclides, based on the detection of annihilation photons formed by the interaction of emitted positrons with electrons in the tissue of a patient, to whom a positron radionuclide was applied. During the radioactive conversion of a positron radionuclide, a positron ("positive electron") is emitted from the nucleus. In the material environment, the positron gradually loses energy by collisions with the electrons of atomic shells and zigzags change the direction of motion. After braking (thermalization) of the positron e+ during a relatively short path, the interaction with the electron e- their mutual annihilation occurs - transformation of the electron-positron pair into two gamma photons with energies 511keV, which fly apart from the place of annihilation simultaneously in opposite directions, at an angle of 180° *) - see §1.6, passage "Interaction of charged particles - directly ionizing radiation", Fig.1.6.1 down.
*) This applies exactly in the center of gravity reference system of the positron and the electron. The energy of photons 2
x 511keV is a consequence of the law of conservation of energy (resting energy of electron and positron is m0e .c2 = 511keV), the opposite direction of 180° is a consequence of the law of conservation of momentum. In the case of collisions of positrons and electrons of higher energies, the angle of inclination of annihilation photons would differ from 180°. In the material environment, however, the positron and the electron have relatively low velocities at the moment of annihilation, so that the emitted quantums actually fly in almost opposite directions, with a maximum deviation of approx. + -2.5°. The effect of this angular deviation is discussed below in the section "Spatial resolution of PET".
  Furthermore, own annihilation usually precedes the formation of metastable bound electron-positron system positronium. In the case of the so-called orthopositronium, 3 photons
g can also be emitted, with continuous spectrum. This can only be observed with positron radionuclides in a sparse gaseous medium; in a relatively dense tissue environment, this phenomenon is very rare (for details see §1.5, section "Elementary particles and their properties", passage "Positronium").
   The path of the emitted positron in the substance (tissue) is "zigzag" and depends on its energy. The mean reach or range of the positron determines the average distance of annihilation from the point of positron emission, ie from the beta
+ radionuclide position. Positrons from the point of emission can fly isotropically in all directions, so that the points of annihilation can be anywhere inside a sphere with a radius given by the range of the positrons. The average range of positrons thus limits the maximum physically achievable resolution of PET (discussed below in the paragraph "Spatial resolution of PET").
   Positron emission tomography uses coincidence detection of a pair of photons of gamma annihilation radiation (511 keV energy), which arise during the annihilation of a positron
b+ with an electron and fly out of their place of origin in opposite directions - at an angle of 180°. This coincident - simultaneously - detection of a pair of annihilation photons is used for electronic collimation of g radiation and subsequent reconstruction of tomographic images.
   Note: For scintigraphic detection of annihilation radiation, even a classic scintillation camera with a special "heavy" collimator with sufficiently thick septa between the holes can be used in principle. In this mode, however, only one of the pair of photons is always scanned - it is a single-photon planar or tomographic scintigraphy (planar or SPECT). However, the detection efficiency is very low here (only one photon + low transmittance of collimators + low absorption in a thin NaI(T1) crystal) and the images have poor spatial resolution (usually worse than 10 mm) due to coarse collimators. This scintigraphy is no longer used.
   Some alternatives, such as the multidetector and Compton cameras mentioned above, are still in the laboratory experiment stage and can only be used for scintigraphic imaging of small objects.
Development of PET 
The basic primary particles used in PET, positrons, predicted by P.Dirac in 1928, were first discovered by C.Anderson in 1932 in cosmic rays
(§1.5, part "Elementary particles and their properties"). Coincidence detection of pairs of annihilation quantum radiation from positron radionuclides for gamma imaging was first tested by W.Sweet and G.Brownel in their two-detector motion scanner in the late 1950s, other PET experimental devices were designed at Univ. of Pennsylvania, the first ring detectors designed by R.Robertson and Z.H.Cho. A significant impetus for the development of PET was the synthesis of 18-FDG (fluorine-18 labeled glucose) in 1970 and the discovery of its accumulation in tumor tissues. In the early 1990s, PET gammagraphy began to be used clinically in large laboratory centers, and after 2000 it spread more and more rapidly to clinical workplaces of nuclear medicine. After 2005, most PET cameras are produced in a hybrid combination with X-ray CT imaging - PET/CT (later sometimes also with magnetic resonance PET/MRI) - for advantages see §4.6, part "Hybrid tomographic systems". All complex oncology centers are gradually being equipped with PET/CT devices.

Coincidence detection ® electronic collimation of g- radiation
The photons g generated during e+ e- annihilation have three significant geometric properties :
¨
They fly out of the annihilation site simultaneously and in the opposite direction - at an angle of 180° ;
¨
They move along straight paths ;
¨
They move at a speed of light of 300000 km/s, so they can be detected at laboratory scales practically simultaneously .
   These properties enable the so-called concident detection of pairs of annihilation photons: we place the measured positron emitter between two detectors (small enough in size), the outputs of which are connected to an electronic coincidence circuit. Only pulses corresponding to the simultaneous detection of photons in both detectors pass through this circuit into the evaluation electronic apparatus. Due to the above mentioned geometric properties, only photons from annihilations that occurred on a straight line connection of the sensitive points of both opposing detectors, can be detected in this way. If annihilation occurs outside this straight connecting line, then even in the case of detection of one of the photons by one detector, the other of the annihilation photons is not captured by the opposite detector - the pulse does not appear at the output of the coincidence circuit. Thus, when a pulse appears at the output of the coincidence circuit, it means that e
+ e- annihilation has occurred at some point on the junction of the two detectors.
   If we surround the investigated object with a positron radionuclide by a larger number of oppositely placed detectors in a coincidence circuit, we achieve targeted directional detection of annihilation
g- photons - their electronic collimation, without the need for physical shielding with a lead hole collimator.


Fig.4.3.5. Principle of scanning and reconstruction of positron emission tomography.
Left: Coincident acquisition of annihilation photons
g. Middle: Image reconstruction. Right: Scintiblock with pixeled BGO/LSO crystal and 4 photomultipliers (manufactured by Hamamatsu) .

PET scanning principle
The PET scanning principle is schematically shown in Fig.4.3.5. The PET scintillation camera detector has an annular arrangement of segments of a large number of small scintillation crystals in optical contact with photomultipliers *), which detect flashes caused by the interaction of radiation
g. Due to the relatively high energy of annihilation radiation g 511keV, BGO or LSO material with higher density and higher detection efficiency in the area of higher energies g is used in scintillation crystals, instead of the usual NaI(Tl) - see §2.4. "Scintillation detection and spectrometry", section "Scintillators and their properties". The diameter of the detector ring is usually 60-80 cm.
*) Individual scintillation crystals
(cut into pixels with dimensions around 4x4mm) are fixed in scintiblocks (Fig.4.3.5 on the right) together with photomultipliers, it is described below.
   The investigated object W, in which the b+ -radioactive substance is distributed, is located inside the detection ring of the PET camera (Fig.4.3.5 on the left). If a radioactive b+ -transformation of the radioindicator nucleus occurs at a certain point, the radiated positron e+ after practically 1-3 mm (depending on its kinetic energy *) movement in the tissue by ionization braking practically stops and when interacting with the electron e- annihilates: e+ + e- ® 2 g, where both quantums of annihilation radiation g1 and g2 with energy 511keV will fly away in opposite directions (ie at an angle of 180°), pass through the tissue and are coincidentally registered by an annular scintillation detector in two places (angles j1 and j2 , in the picture marked: j1 ® x1 , j2 ® y1). The sensing ring of the detectors located around the object to be examined thus detects those photons, which have fallen at the same time on the opposite points of the ring. The connection of these places, the so-called coincidence line or response line, passes through the point where e+ e- annihilation occurred. The set of these coincidence lines from individual pairs of detected annihilation photons (xi , yi) then serves to reconstruct the image of the distribution of the positron radionuclide in the investigated object - in Fig.4.3.5 on the right.
*) This range of positron radiation in the tissue determines the basic limit below which it cannot be reached with the resolution of PET imaging. For the most commonly used 18F, the range of positrons in the tissue is about 0.9 mm, which is substantially less than the actual resolution of the PET apparatus. It is discussed in more detail below in the section "Adverse effects of PET".
Differences between PET versus planar and SPECT scintigraphy 
The main difference from conventional planar or SPECT scintigraphy is that PET detectors are not equipped with lead collimators with many holes, as collimation is performed electronically, which leads to significantly higher detection efficiency of PET compared to SPECT (where most of the radiation is absorbed in the collimator septa). Another difference is that the imaging detector of the SPECT camera must rotate around the examined object (patient) in order to store partial projections at different angles
(this is the case with existing SPECT cameras; with newly developed stationary SPECT cameras , acquisition from all projection angles takes place simultaneously - as with PET). For PET, the detectors do not rotate around the patient, they are stationary - ring detectors store data from all projection angles simultaneously. The resulting image can then be reconstructed continuously during the acquisition.
Coincidence PET with double-headed rotating cameras for SPECT 
In the mid-1990s, some scintillation camera manufacturers (first Adac , then Picker , Elscint , GE and others) developed special electronic circuits that allowed to perform positron emission tomography on conventional two- (or 3) - headed cameras used for SPECT. Both heads, placed opposite each other and without collimators, rotated around the object under investigation as in the acquisition of SPECT, but the pulses were fed to a special coincidence device which recorded and evaluated pulses corresponding to the simultaneous detection of 511 kV annihilation photons by both opposing detectors. The software of the evaluation device then performed the reconstruction of the transverse sections in the same way as for PET.
  This solution initially seemed very promising, as it would allow to perform PET even in workplaces that do not have expensive single-purpose equipment, a universal double-headed SPECT camera supplemented with suitable electronics would be enough, possibly using a thicker scintillation crystal. However, experience from practical use has shown that this is a sub-optimal solution, which with its poorer detection efficiency and resolution, it cannot compete at all with single-purpose PET cameras with a ring detector. Therefore, the manufacturers of scintillation cameras soon withdrew from this solution and offer separately classic double-headed cameras for SPECT and separately PET cameras with a ring arrangement of detectors.

Three types of coincidences in PET
In the coincidence detection of annihilation photons, there can be basically three cases where two photons
g are detected simultaneously :
¨ True coincidences
- direct detection of pairs of photons always coming from one e
+ e- annihilation. The annihilation site is located exactly on the line between the opposite detectors, which during the reconstruction creates an image of the radio indicator distribution. For not very high frequencies (count rate) the number of true coincidences increases practically linearly with activity in the field of view, at higher frequencies it grows more slowly due to dead time and at very high frequencies it even decreases due to overload by random coincidences (paralyzable dead time effect).
¨ Scattering coincidences
- one or both simultaneously detected photons succumbed to Compton scattering, which deviated their angle. The annihilation site does not lie on the junction of the detectors that registered this pair of photons. The percentage of scattering coincidences increases with the (electron) density of the material environment and their number again increases essentially linearly with activity in the field of view
(similar to true coincidences) . If only one of the annihilation photons hits one of the detectors and the other escapes outside the opposite detector after Compton scattering, no coincidence is recorded.
¨ Random coincidences
- this is the detection of photons
g originating from different annihilations, which accidentally hit the opposite detectors simultaneously (within the time resolution of coincidence). The location of neither of this twoo annihilation do not lies on the junction of the detectors that registered it. The number of random coincidences is proportional to the square of the activity of the positron emitter in the field of view.
   Only true coincidences produce a correct gammagraphic picture of the positron radionuclide distribution. Scattering and random coincidences are parasitic (the relevant coincidence lines are false, they do not reflect the actual distribution of the positron radioindicator - this is the so-called combinatorial background) and degrade image quality - reduce contrast and increase noise.

Newer types of PET cameras consist of several coaxial rings of detectors arranged side by side, which allows the simultaneous scanning of several transaxial sections; the field of view in the axial direction is approx. 15 cm for current devices. In this arrangement, two types of scanning are used :
   In the so-called 2D method, shielding baffles are inserted between the individual detection rings, so that the coincidence lines are scanned separately from each cross section - only in the plane of the rings, perpendicular to the system axis.
   In the so-called 3D method, no septa are inserted between the detector rings, and coincidence sensing also takes place "obliquely" from the directions between the planes of the individual rings - the coincidences from the detectors in the different rings are also evaluated. Thus, significantly more photons can be captured, ie achieved higher sensitivity. However, there is also an increased probability of accidental coincidences (see below), so this method can only be fully utilized with cameras with faster detectors based on LSO scintillators.
  For imaging larger parts of the body or for full-body imaging, PET cameras are equipped with an examination bed with a motorized controlled movement. The computer system then combines the scanned data from several patient positions during reconstruction into one large whole-body tomographic image.

Scintillation Detectors for PET
Scintillation Crystals 
As mentioned above, at a relatively high energy of 511 keV annihilation gamma radiation, conventional NaI(T1) scintillation crystals have low detection efficiency. Higher density scintillation materials are more suitable for PET to achieve high detection efficiency with not too large a crystal thickness - to achieve high detection efficiency and good spatial resolution of scintillation localization by a photomultiplier system in the camera's ring detector. It is also highly desirable to have a short scintillation duration (scintillation afterglow) so that a narrow coincidence window can be used - high time resolution to reduce random coincidences
(and the possibility of using the TOF method - see below). To detect gamma radiation, a number of scintillation materials with different properties have been synthesized (see §2.4, section "Scintillators and their properties"). In principle, several types of scintillators are applicable for PET :

Scintillator : NaI (Tl) BaF2 LaBr3 (Ce) YAlO3 (Ce)
(YAP)
LuPO4 (Ce)
(LPO)
Gd2 SiO5 (Ce)
(GdSO)
Bi4 Ge3 O12
(BGO)
Lu2 YSiO5 (Ce)
(LYSO)
Lu2 SiO5 (Ce)
(LSO)
Lu Fine Silicate
(LFS)
LuAlO3 (Ce)
(LuAP)
Density [g/cm3] 3.67 4.89 5.1 5.55 6.2 6.71 7.13 7.1-7.4 7.41 7.35 8.34
lmax [ nm ] 415 220/310 360 350 360 440 480 420 420 425 380
scint. afterglow [ns] 230 0.8 16 30 24 60 300 41 40 33 11/28
h [photon/MeV] 4.10 4 1.8.10 3 6.3.10 4 1,6.10 4 1,3.10 4 8.10 3 6.10 3 3.10 4 3.10 4 3.2.10 4 9,6.10 3

In practice, heavier BGO scintillators are used, more recently LSO (possibly LYSO modified), which has the advantage of a significantly shorter scintillation afterglow. LYSO scintillators have similar properties to LSO; the yttrium component causes technologically easier growth of single crystals. A higher percentage of yttrium (LYSO also occurs on the composition of Lu0.6 Y1.4 SiO5 : Ce), but reduces the density and the detection efficiency in comparison with the LSO.
   Based on LSO, the LFS (Lutetium Fine Silicate) scintillator was further developed, which has a finer crystal structure and in addition to basic lutetium, silicon, and oxygen (LSO) with doping Ce, it also contains carefully tested small impurities of some other elements such as Ca, Gd, Sc, Y, La, Eu, or Tb. This results in slightly better energy resolution and shorter scintillation afterglow.

  Scintillator LaBr
3: Ce5% (is hygroscopic) with a very fast scintillation is tested in terms of TOF (see below).
Internal radioactivity of LSO scintillators 
A minor disadvantage of lutetium -based scintillation detectors (such as LSO and LYSO) is the higher radiation background due to the internal radioactivity contained in the scintillator. In addition to the basic stable isotope
175Lu, natural lutetium also contains 2.6% of the long- lived radioisotope 176Lu with a half-life of 3.8.1010 years is also contained in the luterium - see §1.4, passage "Lutetium". This natural "contaminant" is irremovable. During its radioactive decay, beta and gamma radiation is emitted, which is internally detected with high efficiency and causes an internal radiation background in each detector *) about 40 pulses/sec / l gram LSO (more detailed analysis was performed in §2.4, part "Scintillators and their properties", passage "Internal radioactivity of LSO scintillators"). Beta radiation is fully absorbed in each individual crystal independently, so it does not manifest itself in coincidence measurements (random coincidences are negligible here). However, gamma-ray beams, especially 300 keV photons, can fly out of individual LSO crystals and hit other detectors, where they can be detected immediately - creating a coincidence event contributing to the background in the PET image. The background thus formed is negligibly small in relation to the fluxes of the measured annihilation radiation of the order of 106 photons/s. in clinical scintigraphy. However, certain problems may arise in experimental studies of PET with low activities of the order of kBq units at long measurement times (animal PET).
*) It is interesting that a typical PET camera, consisting of about 190 blocks of LSO crystals with a volume of about 50 cm3, contains a total internal radioactivity of 176Lu of about 2.4 MBq! Each 50 cm3 scintillation detector produces about 12,500 pulses/s. radiation background, which significantly burdens the electronic reading circuits. However, when coincidently measuring higher activities (approximately 100 MBq in patients), this is practically not applied in the resulting images. However, this "parasitic" radiation can be used for continuous calibration and tuning of PET detectors, without the use of external phantom sources..?..
Photodetectors for PET 
Two types of photodetectors are used in PET to capture and electronically register light flashes from scintillation detectors :
× Photomultipliers 
are the most commonly used and proven electronic light signal sensors from scintillation detectors - they are described in detail in §2.4, section "
Photomultipliers". They have high and linear gain, low signal-to-noise ratio, short output signal pulse (short dead time). Their partial disadvantages are the complexity of the design, the need for high voltage, larger dimensions (they cannot be miniaturized too much), higher cost, relatively low quantum efficiency and sensitivity to the magnetic field.
×
Semiconductor detectors
are a modern alternative to photomultipliers. Their main advantages are: small compact dimensions (miniaturization), high quantum efficiency, low voltage, lower cost, insensitivity to magnetic fields. Two types of semiconductor photodetectors are used for scintillation sensing :
- Photodiodes are formed by p- and n-type semiconductors in close contact in the p-n junction. They are connected in inverse polarity to voltage (in the reverse direction). The impact of the photon of light excites electron-hole pairs in the semiconductor material, whereby a current pulse passes through the diode. At higher voltages, secondary electron-hole pairs also form and signal amplification occurs. If the electric voltage is set just around the breakdown voltage of the p-n junction, there will be an avalanche-like increase in electron hole pairs when the photon strikes - it is avalanche photodiode, operating in the so-called Geiger mode.
- Silicon "photomultipliers" SiPM are multipixel avalanche photodiodes - multipixel photon counters, each element of which works independently in Geiger mode. The output signal is proportional to the number of pixels that have been hit by light photons, and thus the number of photons detected in the flash, they have spectrometric properties (they are described in more detail in §2.4, section "Photomultipliers", section "SPM Semiconductor Photomultipliers").
Detector blocks for PET 
Scintillation crystals with photomultipliers (or semiconductor photodetectors) are assembled into compact scintiblocks in a PET camera, distributed around the circumference of the circular gantry. Each such scintiblock is formed by a square 2D array of crystals (BGO or LSO), connected to photomultipliers by means of a light guide - Fig.4.3.5 on the right. The array of crystals is usually formed from one single-crystal using sections separated with light-reflecting material. The usual configuration consists of a crystal measuring 5
x5 cm and 3-5 cm thick, cut into an array of 8x8 partial crystals ("pixels"), to which 4 photomultipliers with a diameter of approx. 2 cm are attached via a light guide (in Fig.4.3.5 cutout at the top in the middle, in more detail in the right part of the picture). When the photon g of radiation hits one of the crystals, the resulting scintillation light is shared by all four photomultipliers. Information on the exact position of the flash (x, y coordinates) in the crystal field is obtained by electronic analysis of the ratio of pulse amplitudes at the output of individual photomultipliers, similar to a classical planar gamma camera (described above in §4.2 "Scintillation camera", Fig.4.2.1; each PET scintiblock can be considered a simple "Anger mini-camera"). One ring of the detector is usually formed by made 48 scintiblocks, arranged close together in a circle with a diameter of about 60-70 cm (gantry); the whole camera contain 3-5 such parallel rings.
   As mentioned above, instead of photomultipliers, the multicrystal scintillation can electronically scaned using arrays of semiconductor photodiodes, or better with SiPM photomultipliers. The near future probably belongs to compact scintiblocks LSO-SiPM, LYSO-SiPM
(or perphas LaBr3 -SiPM). For more distant development in PET (instead of BGO/LSO scintiblocks with classical photomultipliers or SiPm) there are promising multipixel fully semiconductor detectors (eg based on CZT). In addition to better detection efficiency and spatial resolution, a slightly shorter coincidence time (for better TOF) can be achieved. So far, it is being tested experimentally on smaller PET models. The advantage of semiconductor detectors is also theirs independence from the magnetic field, which allows use also in hybrid PET/MR systems.

Reconstruction of PET images
During the acquisition, a large number of coordinates of coincidence lines (in the order of millions of coincidence detections) are scanned; data are stored in the form of so-called sinograms. By computer reconstruction of these linear projections of coincidence sites, images of cross-sections are created and from a set of transverse sections, computer reorientation can be used to create sections at any angle, or 3D images (similar to SPECT, above). For the reconstruction, either the (filtered) back projection method is uses, which however, can produce "star" artifacts around positive lesions, or more computationally demanding iterative reconstruction, providing higher quality images without these artifacts. Another advantage of iterative reconstruction is the ability to incorporate various properties of specific devices and methods (such as homogeneity, attenuation, noise, resolution) directly into the reconstruction procedure. Reconstruction methods, analogous to SPECT, have been described above in the section "
Computer reconstruction of SPECT".
................? add special modifications of reconstruction procedures? ........

PET imaging properties
Compared to classical single-photon planar and SPECT scintigraphy, two-photon coincidence tomography of PET has two basic advantages: significantly higher detection efficiency (sensitivity) and slightly better spatial resolution :
Detection efficiency (sensitivity) of PET 
The absence of classical collimators and registration of photons of annihilation radiation simultaneously from all directions, using electronic collimation, leads to significantly higher detection efficiency (sensitivity) of PET gamma cameras compared to classical Anger cameras, where the vast majority of gamma photons are not detected
(flies "into empty space" or is absorbed in the septa of collimators).
   The detection efficiency or sensitivity
h of instrument for detection and spectrometry of ionizing radiation is generally a ratio between the number of detected pulses and the number of incoming radiation quanta; relative and absolute efficiency is introduced, often expressed in % (it was defined and physically discussed in §2.1, section "General physical and instrumental effects in detection and spectrometry", section "Detection efficiency and sensitivity"). However, in gamma cameras, where the source of gamma radiation is a radionuclide, the sensitivity - detection efficiency - is usually quantified in a special way: as the number of pulses detected by the camera per unit time (per second) - [cps], relative on unit of activity [kBq, MBq] of the radionuclide used in the displayed source; for the planar/SPECT scintigraphy is usually 99mTc, for PET it is 18F. Only exceptionally is it expressed in % (detection efficiency-sensitivity in classical gamma cameras is discussed in §4.5, section "Sensitivity ( detection efficiency ) of a scintillation camera"). For the scintillation cameras, the detection sensitivity is related to the radioactivity of the examined object: the detection efficiency h, or sensitivity, of the imaging system is quantified as the pulse frequency N[imp./s] measured by a scintillation camera with a (point) radiation source g (located at the desired place in field of view), relative to the activity unit A[MBq] of the source: h = N / A. It is expressed in units [imp. s-1 MBq-1], or [cps/MBq] or [cps/kBq].
   Detection efficiency - sensitivity - of PET cameras are determined by several physical, geometric and technical factors, which can be divided into two categories :
l Detection efficiency hd of annihilation radiation 511keV in the detection elements of the PET ring. This "physical" detection efficiency hd of annular detectors depends on the thickness of the detection material h , its density and the atomic number - trough the linear attenuation factor m for gamma 511keV is given by the coefficient (1-e- m . h) (detection efficiency of scintillation detectors is discussed in §2.4 "Scintillation detection and gamma radiation spectrometry", section "Scintillators and their properties"; scintillation materials suitable for gamma 511keV were discussed above in the section "Scintillation detectors for PET"). It also depends on the amplitude analyzer window setting, scintillation afterglow and dead time. At greater distances from the center, there is also a slight decrease in the detection efficiency due to the oblique angle the impact of most of the annihilation photons on the detectors. We include all these other individual influences in the factor f . In our case of coincident two-photon detection of PET in two opposite detectors with detection efficiency hd, the resulting detection efficiency will be given by the product hd . hd , ie it appears in the quadrate hd 2 = [f . (1 - e - m .h)] 2.
l Geometric efficiency hg the PET registration of 511 kV annihilation photons is given by the spatial angle of projection at which the annihilation photons from the activity source are detected. Each radionuclide source emits radiation isotropically in all directions - up to a spatial angle of 4p. Around the point source, located in the middle of the detection ring of radius R, we can draw an imaginary spherical surface of this radius R - trough its surface 4pR2 will pass all photons emitted from the source (if the detectors were placed densely on this spherical surface, the geometric detection efficiency would be 100 %). By circular detection ring width S, which has a surface 2p R.S, however, passes out of this total number only a part of the photons, given by the area ratio of 2p R.S/4p R2 = S/2R. If the PET camera has N parallel rings of width S and radius R, the geometric efficiency will be hg = N.S / 2R (if we neglect the gaps between the detectors and somewhat oblique angles from the center to the peripheral rings). Increasing the diameter of the ring 2R reduces the overall projection solid angle and thus the geometric efficiency. On the contrary, with the increasing number of N detection rings in a PET camera, the projection angle widening and thus increases also the geometric detection efficiency.
   By multiplying these two partial factors of detection and geometric efficiency, we can obtain the resulting relationship for the total detection sensitivity h of the PET camera :
          
h  =  [f . (1 - e - m . H )]2 . N . S / 2R  ,
where f is the fraction of detected photons in the photopeak with the specific setting of the PHA analyzer window
(with possible angular dependence in the peripheral parts), h is the thickness of the detectors, m is the linear attenuation factor of the detector material used for gamma radiation 511keV (usually BGO or LSO m = ....) , R is the radius of the detection ring, S is its width, N is the number of detection rings.
   Current standard PET cameras with three detection rings with a diameter of about 70 cm, each consisting of 48 scintiblocks BGO/LSO 5
x 5 cm and a thickness of about 5 cm, achieve a detection sensitivity of about 7-10 cps /kBq 18F (for classic Anger planar or SPECT cameras sensitivity is only about 0.04-0.3 cps /kBq 99mTc, depending on the collimator used - almost 30-100 times lower).
Spatial resolution of PET   
Replacement of mechanical collimators by electronic collimation also leads to a somewhat better spatial resolution of PET compared to conventional Anger cameras.

   The spatial resolution (abbreviated resolution) of a scintigraphic image
is the smallest distance [mm] of two point radioactive sources in the displayed object, which are still distinguishable from each other in the scintigraphic image as two images. We can determine it as the width of the PSF profile in the image of a point or line source in half the maximum height of the profile, converted to a spatial scale in the object [mm] - it is called FWHM ( Full Width at Half Maximum - overall width at half maximum; resolution for classical gamma cameras is discussed in §4.5, passage "Spatial resolution") .

Fig.4.3.6. Physical and geometric effects on positional resolution in PET imaging.
a) Range of positrons h in the tissue (an example of three ranges in different directions is marked - strongly increased) and flight deviation of annihilation photons from 180°. b) Projection-geometric degradation of resolution due to width d of detectors. c) Cros-radiation of annihilation photons between detection elements causes radial "astigmatism" in images of peripheral parts.

The spatial resolution of PET imaging is again determined by a combination of several physical, geometric, and technical factors :
¨
The range of positrons in the tissue prior to positron annihilation represents the primary physical limit for PET resolution. Positrons are emitted from nuclei at high speed - with a kinetic energy of hundreds of keV to several MeV, so from the site of radionuclide deposition positrons to fly in the tissue to a certain distance (approx. 0.5 mm-6 mm, depending on the radionuclide) before braking (thermalizing) and meeting with electrons (via positronium) annihilate on a pair of 511keV photons. Therefore, since the positions in which annihilation photons are generated are somewhat different (and each time different) from the position of the original parent nuclei, there is some blurring of position. The magnitude of this blur depends on the positron radionuclide used - on the maximum and mainly the mean energy of the emitted positrons, determining their mean range h in the tissue (Fig.4.3.6a). For some positron radionuclides used, the following is :
....... table ..............
¨ Angular blurring due to incomplete braking of positrons. Electron-positron annihilation in the tissue then occurs with a certain residual kinetic energy (different each time), so that in the laboratory reference system the angle of radiation of annihilation photons will be slightly different from 180°, on average by + - 0.25°. This angular uncertainty - variability causes a small geometric blur, which is proportional to the radius R of the detection ring, with a value of 0.004 . R .
¨ Size of detectors - width d of detection elements is the dominant factor of projection-geometric degradation of resolution. Opposite detectors of finite (non-zero) dimensions detect radiation not only from a single (central, axial) coincidence lines, but from the whole cone of angles. This leads to a geometric variability of the coincidence line with respect to the actual position of the annihilation, the half-width of which is d/2 (Fig.4.3.6b).
¨ Accuracy of decoding the position of scintillation within the scintiblock of detection elements, using a significantly lower number of photodetectors (eg 4 photomultipliers per 64 detection scintillation elements). The inaccuracy of this decoding (optical multiplexing) somewhat degrades the resolution. For this contribution, the value of the half-width of approx. d/3 was empirically determined.
¨
Cross-radiation of annihilation photons between detection elements - penetrating annihilation g of 511keV radiation can interact with several different (adjacent) crystals. This can cause a detected signal even in another neighboring crystal than the one on which the photon primarily strikes. As a result, the coincidence line may be incorrectly assigned to one of the adjacent detection elements. This effect occurs when annihilation photons hit the detection elements in an oblique direction, which occurs from sources farther from the center (for sources in the center r = 0 it does not manifest itself and the projection of coincidence lines remains narrow, given only the size of the detection element) - Fig.4.3.6c. These obliquely incident photons can interact with several different crystals, depending on the depth of penetration into the scintillation material. The radial projection of the source is thus extended by the trigonometric factor k . r / Ö(r2 + R2), wherein the coefficients k a given depth of penetration annihilation photons into scintillation material- the half-layer of photons 511 keV absorption in the material of the detector h1/2 = ln2 / m , where m is the linear attenuation coefficient gamma 511 keV in the material scintillator - Fig.4.3.6c.
   This creates a kind of radial "astigmatism" *) - asymmetric blurring of images of peripheral sources, more distant from the center of the ring r = 0. For detection crystals made from BGO or LSO scintillators, an approximate value of 12.5 was measured for the coefficient k. This effect is clearly visible in Fig.4.3.7 on the right, in images and profiles of peripheral point sources at distances r = 34 and 30 cm, partly also for r = 20 cm
(allusively already manifests at r = 10 cm).
*) It is a bit similar to astigmatism in optics - imaging defect (aberration) in converging lenses, manifested in objects at greater distances from the optical axis, or in optical systems asymmetrical to the optical axis.
¨ Reconstruction algorithms for creating the resulting images of radioindicator distribution using a set of coincidence lines show minor errors and variations. Manifests a non-uniformity of the density of coincidence lines with respect to the position of sources inside the detection ring, different types of reconstruction algorithms and filtering. During the reconstruction, a certain common additional coefficient of degradation of resolution *) is arises, for which the range of approx. 1.2-1.5 was empirically determined; we will use an approximate value of 1.3 here.
*) In this physical analysis we mean standard "classical" reconstruction algorithms of filtered back projection or iterative reconstruction of OSEM. We do not consider special reconstruction algorithms with built-in resolution recovery, design modifications and PSF modeling or filtering using inverse MTF to artificially improve resolution. Here we are interested in the purely physical properties of PET images, not the possibility of their additional computer improvement (which, however, can be useful in practice...).
   These effects lead to several contributions to the response function of the PSF point source, which are approximately Gaussian in shape. By their quadratic summation (geometric average) we can then obtain the resulting relationship for the total spatial resolution of the PET image :
          FWHM  =  1,3 .
Ö [(d/2)2 + h2 + (0,004.R)2 + (d/3)2 + (12,5.r)2/(r2+R2)]   ,
where FWHM is the resulting half-width of the response function of the point source, ie the total resolution, d is the size (width) of the detection element, h is the mean range of the positrons, R is the radius of the detector ring of the PET camera, r is the radial distance of the source emitter from the center of the ring
(all dimensions are in millimeters).
   Current standard PET cameras with three detection rings about 70 cm in diameter, each consisting of 48 scintiblocks 3 cm thick with detection elements about 5
x5mm in size, achieve spatial resolution in the middle of the ring in transverse sections around 4.2-5 mm (classic planar/SPECT cameras they reach such a resolution only close to the front of the collimator, in practical scintigraphy, where the distance - depth - of the lesion is around 10 cm, but the resolution cannot be achieved better than 10-12 mm).
Small animal PET cameras with a diameter of approx. 20 cm with detection elements with a width of approx. 0.5-1 mm achieve an even significantly better resolution of approx. 1-1.5 mm .

Fig.4.3.7. PET images of point sources 18F located at different distances r from the center of the detection ring. By analyzing the ROI and profile curves with these images, the values of detection efficiency h and spatial resolution FWHM were measured (we measured on a PET camera GE Discovery IQ at KNM FN Ostrava) .
At the last peripheral point source at a distance of r = 34 cm, part of its image was already cut off by the edge of the field of view.
The measurement was performed using a simple arrangement described in the work "Phantoms and phantom measurements", part "
Tomographic phantoms", passage "Simple improvised phantom for measuring the imaging properties of a PET camera".

Our measurement of spatial resolution and detection sensitivity on PET images of point sources, located at different distances from the center of the detection ring, is marked in Fig.4.3.7. The effect of radial astigmatism can be clearly seen in the images and profiles of peripheral point sources at distances r = 34 and 30 cm, partly also for r = 20 cm (allusively are visible already at r = 10). For similar reasons (due to the oblique angle of incidence of most of the annihilation photons on the detectors) there is also a slight decrease in the detection efficiency at greater distances from the center.

TOF - time localization of the annihilation site
The basic (conventional) PET method described above does not give any information about the annihilation site on the coincidence line, all pixels on the coincidence line are assigned the same annihilation probability, the image is formed only by intersections of coincidence lines. However, increasing the speed of electronics and the introduction of detectors with high time resolution
(such as LSO scintillators) gradually allows the use of another important "information channel" of annihilation radiation for PET: it is measuring so-called TOF (Time Of Flight) - flight time of photons g from the annihilation site to the detectors. These photons fly in opposite directions at the speed of light c = 300000 km/s. If annihilation occurs in the middle of the coincidence line "0", both photons are detected exactly at the same time. However, if an annihilation occurred off-center, at a distance Dx, the photon g1 will have a flight time TOF1 = x1/c to the detector, while the second photon g2 will fly a slightly different time TOF2 = x2/c - see Fig.4.3.8. From the time difference t2 - t1 = TOF2 -TOF1 it is then possible to determine the radial coordinate Dx of the annihilation site on the coincidence line: Dx = c. (t2 - t1) /2 .

Fig.4.3.8
Time localization of the annihilation site x
1, x2 on the coincidence line by electronic analysis of the difference in flight times of annihilation photons DTOF in positron emission tomography.

If the coincidence detection of annihilation radiation has a sufficiently short time resolution, the time difference between the detection of both annihilation quanta g can be measured, which allows (at least in principle) to determine the place on the coincidence line Dx, where annihilation occurred and from where both photons were emitted * ) - Fig.4.3.8. This introduces additional information about the position of the detector response into the system.
*) If we could measure the time differences of annihilation photon arrivals with picosecond accuracy, this information would be enough to determine the sites of annihilation and achieve PET imaging directly. There would then be no need to reconstruct by the quantifications of the intersections of the coincident rays, but the cross-sectional image could be stored directly (in polar coordinates). These would no longer be coincidence lines, but coincidence points. However, for it we do not yet have fast enough electronics and detection technology...
   The time resolution of existing instruments does not yet allow accurate localization of annihilation sites, but even the approximate location of the annihilation photon radiation site could shorten the segment of response line, improve the reconstruction procedure, and improve the signal-to-noise ratio in the resulting images. A general analysis of the SNR signal-to-noise ratio in scintigraphic images and its effect on lesion recognizability was performed above in the section "
Scintigraphic image quality - lesion recognizability". There it has been shown that this ratio SNR = (A-B)/ÖB, where A is the number of useful pulses and B is the number of background pulses. If at PET we have the total number of background pulses BD on the whole concidence line of length D (given by the diameter D = 2.R of the detector ring), then in its portion FWHMTOF , selected by the width of the TOF length resolution window, this number of background pulses will be reduced to BTOF=BD.(FWHMTOF/D), while the number of useful pulses A remaining the same; this applies to all coincidence lines. Thus, the use of TOF localization of the annihilation site on the coincidence line leads to an improvement in the signal-to-noise ratio by a factor of SNRD/SNRTOF = [D/FWHMTOF]1/2.
   TOF has no direct effect on improving the spatial resolution FWHM of the PET camera, if the PET image is reconstructed using the intersections of the coincidence lines - this follows from the above geometric analysis according to Fig.4.3.6. Only if the length resolution of FWHM
TOF could be reduced to a few millimeters, the reconstruction of PET could take place with significant use of the parameter Dx according to Fig.4.3.8, and TOF would thus also affect the spatial resolution of PET (in the extreme case of a perfect TOF, its length resolution could even be decisive for the resulting spatial resolution of PET - however, this is not "threatening" yet..!..).
   TOF is still in the stage of technical development. The initial enthusiasm has not yet been fulfilled, the method is rather a promise for the future for the next generation of PET cameras. For current types of PET cameras (2010-20) with installed TOF, the TOF time resolution is about 500-600 picoseconds, which corresponds to the possibility of resolving the annihilation site on the coincidence line of about 15-18 cm. The TOF parameter is so far only minimally usable in clinical scintigraphic diagnostics. TOF analysis will be relevant only when its length resolution from the current 15 cm can be improved to about 5-2 cm...
   An improvement in TOF time resolution to less than 400 ps can be expected with the introduction of special scintillators
(such as lanthanum bromide LaBr3) and modern photodetectors (silicon photomultipliers SiPM). Data are loaded and reconstructed in LIST-mode format, iterative reconstruction procedures (3D list mode TOF MLEM, OSEM, ...) have built-in special correction algorithms, containing data from calibration measurements. So far, however, "no miracles" can be expected.!.. - TOF will probably remain mostly a physical-technical interest for a long time...

Adverse effects at PET and their correction
As with planar and SPECT scintigraphy, there are some adverse and disturbing effects also on PET imaging, which worsen the quality of the images. We will mention here several significant adverse effects, of which the first four are also known from planar and SPECT scintigraphy, the others are specific for two-photon PET :

The same applies to the pitfalls and possible errors of correction methods, as in the section "Errors and pitfalls of correction methods - correction artifacts" in general scintigraphy.

Construction design of PET gamma cameras
The basis of each PET gamma camera
(for which there is a less suitable name "PET scaner") is a circular ring of detectors with a diameter of 60-80 cm, registering pairs of photons of annihilation gamma radiation with energy 511keV in coincidence mode from opposite directions. The ring consists of a large number of detectors (mostly scintillation), whose scintiblocks (see Fig.4.3.5 on the right) are mounted in several parallel concentric rows on a gantry, through the inner cylindrical "tunnel" of which moves the lounger with the patient. The movement of the lounger is driven by a servomotor, ensuring precise computer-controlled movement so that the images captured by the rings from the individual parts of the body are folded into the resulting PET image (sometimes even whole-body), including online fusion with X-ray CT images.

Fig.4.3.9. PET/CT positron emission tomography
examination room at the Department of Nuclear Medicine, University Hospital Ostrava.
   In the middle is the basic PET / CT device
(GE Discovery). Aiming and navigating laser pointers for radiotherapy planning are installed on the sides and ceiling of the laboratory. In the back of the right there is a contrast agents applicator for CT on the stand.

Current PET devices are two-modality - PET/CT to assess the exact anatomical location of imaged lesions by fusion with CT X-ray images (or PET/MRI for fusion with nuclear magnetic resonance images). In addition to the PET ring, a CT ring (or MRI) is also installed on the same gantry, through which the bed passes in a controlled manner and simultaneously with the PET imaging, it also in-line creates CT images (or MRI) of the patient. Beside the physiological-anatomical correlation, CT images also provide density maps for the correction of attenuation of gamma annihilation radiation in tissues at PET detection.
   The device sometimes includes optical aiming and navigation lasers for precise localization of lesions when defining ROI within the irradiation plan using PET images.

Use of PET scintigraphy
The areas of clinical use of positron emission tomography in nuclear medicine are intended, similarly to emission planar and SPECT scintigraphy, mainly by the properties of relevant radiopharmaceuticals, here radiopharmaceuticals labeled with positron radionuclides
(these radionuclides and radiopharmaceuticals are briefly described in §4.8 "Radionuclides and for scintigraphy"). The most important area of PET use is oncological diagnostics - finding out the location and nature of tumors, which accounts for more than 90% of all PET examinations (§4.9.6 "Scintigraphic diagnostics in oncology" and §3.6, section "Diagnosis of cancer"). To assess the precise anatomical localization of the displayed lesions, is used fusion of PET with X-ray CT images - twoo-modality PET/CT, or with MRI images (PET/MRI).

 

Fig.4.3.10. Example of PET / CT scintigraphy with 18FDG in a patient with lymphoma.
   PET images show multiple foci of increased glucose metabolism in the lymph nodes of the neck, left axilla, mediastinum, retroperitoneum, pelvis and groin, indicating viable tumor neoplasia.
   
Physiologically, 18-FDG is deposited in the brain, in the hollow kidney system, bladder, slightly diffusely in the liver and spleen, accumulation in the myocardium is common.
   The transverse section shows separately a significant deposit in the area of the supraclavicular (indicated by arrows) for the assessment of metabolic accumulation using the SUV value [g/ml.] :
  
The SUVmax of the lesion in the supraclavicular is 11.5, with the reference SUVmax liver 2.9 and the mediastinum 1.4.

(PET / CT images were taken by
Martin Havel, MD, Ph.D.,
head of the PET/CT department
of KNM FN Ostrava )

At a general level, PET works very well in the field of assessing the metabolic activity of tumors, proliferation, tissue hypoxia, the density of expressed receptors in cells.
   Specifically, herein used pharmacokinetic properties especially
18F-deoxyglucose FDG (hereinafter also 18FLT, 18F-choline) which is selectively uptake in tumor cells with increased metabolism of carbohydrates - appears metabolic cellular activity of tissues (whereas X-ray and ultrasound displays only morphological page). Malignant tumors are usually characterized by glucose hypermetabolism. This method is therefore also suitable for monitoring the response of tumor tissue to therapy by imaging metabolically active tumor tissue as opposed to inactivated cells; it is possible to monitor the therapeutic response - the "success" of therapy. Among other things, it is able to recognize tumor recurrence from other processes (eg from the consequences of previous tumor treatment), see §3.6, section "Modulation of radiation beams". Monitoring of the therapeutic response by PET consists in comparing the metabolic activity of the tumor before the start of treatment and after the application of therapy. The change in tumor metabolism occurs before the change in its dimensions, assessed by morphological X-ray or sonographic imaging methods. PET can also be used for detection inflammatory process in the organism (§4.9.6 "Oncological radionuclide diagnostics. Scintigraphy of inflammation"), in cadiology for the diagnosis of myocardial viability (§4.9.4 "Nuclear cardiology").
   The metabolic activity of tumor lesions is often quantified in PET images using SUV values - from a general point of view, it was discussed above in §4.2 "
Quality of scintigraphic imaging", section "Quantification of positive lesions - SUV". There were discussed some physical and biological factors that may skew the absolute SUV quantification and recommended relative SUV quantification to compare the metabolic activity of lesions in specific patients before and after therapy. In the case of PET imaging with 18F-FDG, the blood glucose level also adds to this - its increased value reduces the accumulation of FDG, which underestimates the SUV.
   PET (/CT) has thus become an important functional imaging method in the primary diagnosis, staging, assessment of therapeutic response, recurrence search or re-staging of a number of oncological diseases. PET images can also be advantageously used for radiotherapy planning. 
It is discussed in more detail in §3.6, section "Diagnosis of cancer".
   For PET scintigraphy with the most commonly used
18F-FDG is a problem of physiologically variable (sometimes quite high) accumulation of FDG in a number of healthy viable tissues, especially in the brain and myocardium (see Fig.4.3.10). Therefore, PET with FDG is not suitable for the detection of brain metastases or minor perfusion defects of the myocardium. Non-specific increased accumulation of FDG is also observed in inflammatory processes, wound healing, after surgery. The displayed site with increased glucose uptake may not always be a tumor...
   Therefore, is promissing the use of other radiopharmaceuticals and positron radionuclides
(§4.8, passage "Radionuclides and radiopharmaceuticals for PET") such as gallium 68 Ga, zirconium 89 Zr , iodine 124 I , copper 64 Cu (for PET) and beta- 67 Cu (for therapy), scandium 44 Sc (for PET) and beta- 47 Sc (for therapy), or mixed alpha-beta+ terbium 149 Tb (for both PET and alpha-therapy). These radionuclides can be used to label mainly monoclonal antibodies for PET diagnostics and to perform subsequent biologically targeted radionuclide therapy - theranostics (§4.9, passage "Combination of diagnostics and therapy - teragnostics").
  
An interesting application of PET has recently appeared in the so-called hadron radiotherapy (§3.6 "Radiotherapy", part " Hadron radiotherapy "), where irradiation with high-energy charged particles in the irradiated tissue causes, among other things, nuclear reactions, during which positron radionuclides are formed. When irradiated with accelerated carbon nuclei 12C, a positron radionuclide 11C is also formed, the distribution of which can be visualized by the PET method. With a PET camera installed on a hadron radiotherapy irradiator, we can monitor the dose distribution in the target tissue and in the surroundings - so-called in-beam PET monitoring - and thus control the course of radiotherapy (see Fig. 3.6.6 in §3.6).
  PET is also used in neurology to diagnose brain activity and perfusion. The area of the brain that is active has an increased accumulation of radiopharmaceuticals, which can be used to assess brain activity and its association with some psycho-neurological disorders (including Alzheimer's disease). Furthermore, it is a scintigraphic diagnosis of inflammatory processes and examination of the myocardium, where the perfusion and viability of the myocardium can be assessed on the basis of the consumption of special positron radiopharmaceuticals (see below §4.9.4, section "Myocardial perfusion").
   For clinical applications is very important the fusion combination of PET scintigraphy with X-ray CT imaging (anatomical - §3.2, part "Transmission X-ray tmography CT"), which provides visualization of morphological and anatomical structures with high spatial and density resolution. This information obtained from CT can be used to increase the accuracy of the location, extent, and nature of the lesions found in PET images. X-ray CT thus complements the functional information obtained by PET with the help of a radiopharmaceutical, with localization anatomical information. For modern devices, this is implemented online in a two-mode hybrid PET/CT system - see below §4.6, section "Hybrid tomographic systems". This combination also allows good correction for absorption (attenuation) of the annihilation g radiation in tissue. Recently, there is also a hybrid combination of PET/MRI.
Positron emission mammography (PEM) 
The PET method using suitable tumor-accumulating radiopharmaceuticals (usually 18 FDG or FLT) is naturally also used in the diagnosis of breast cancer. However, the specific anatomical proportions of the breasts and the properties of mammary lesions have led to efforts to develop smaller single-purpose - dedicated, optimized - PET imaging devices that would have a higher resolution for small lesions typical of breast cancer. And also a shorter acquisition time than with whole body PET and the application of lower radio indicator activity. Basically, two technical solutions of these specialized devices for PEM positron emission mammography have been developed :
- The first with its design resembles a classic X-ray mammography with compression, only the X-ray tube and the detector are replaced by two flat PET imaging camera detectors (in Fig.4.3.11 on the left), between which a breast is inserted with suitable compression.
- The second system is an small annular PET detector of circularly arranged scintiblocks (approx. 48), of significantly smaller dimensions than whole-body PET (approx. 20 cm diameter), usually one ring into which uncompressed breast is inserted. The breast hang freely inside the detector located under the lounger with the hole on which the patient lies - Fig.4.3.11 on the right.


Fig.4.3.11. Technical design of PEM positron emission mammography instruments.
Left: Compression PEM mammogram with flat PET detectors. Right: Ring PEM mammogram with loosely inserted breast without compression.

In both cases, coincidence detection of a pair of annihilation photons by opposing detectors is performed, with computer reconstruction (in the PEM ring it creates the cross-sectional images, as in classical PET). The advantage of these optimized PEM devices is better spatial resolution (approx. 2-3 mm), allowing to detect even small lesions in the breast (in case of good accumulation even under 1 cm).
   Although small PEM devices are significantly cheaper than large universal (full body) PET cameras, the positron emission mammography method has not become more widespread (unlike the widely used X-ray mammography - §3.2, section "X-ray mammography"). One of the reasons is a positron radiopharmaceutical with a short half-life, which is difficult to access outside the larger workplace of nuclear medicine. It serves only as additional method to X-ray, sonographic or MRI mammography. However, in order to visualize the more complex extent of the disease, it is still necessary to perform PET imaging on a larger scale, including nodes and potential metastasis. PEM is a specialized peripheral method that is sporadically used mainly in the USA and Japan in some larg complex oncology centers...
  Specialized PEM devices in the world are supplied by only two manufacturers: CMR Naviscan, California, USA and
IHEP - GaoNeng Medical Equipment, Hangzhou, China .

-------------------- minor physical-technical interest -------------------

Neutron Stimulated Emission Computed Tomography ( NSECT)
NSECT (Neutron Stimulated Emission Computed Tomography) is a new (and so far experimental) method of spectroscopic imaging of the concentration of certain elements in an organism using neutron interaction. Unlike conventional emission computed tomography SPECT or PET, gamma radiation is not emitted by radioactive isotopes, but by stable isotopes in which the emission of
g- radiation (characteristic energy) is stimulated by inelastic scattering of fast neutrons, by which the analyzed area is externally irradiated. These stable isotopes may either be a natural part of the tissue under investigation or may be introduced as molecular indicators (similar to contrast agents or radioindicators), eg by a metabolic pathway.


Fig.4.3.12. Principle of neutron stimulated emission computed tomography NSECT.

The analyzed area (sample, tissue) is irradiated with a beam of fast neutrons (energy approx. 7-10 MeV) from a suitable collimated neutron source - electronic neutron generator (§1.5, part "Accelerators", passage "Accelerators as neutron generators") or radioisotope source ( .....). These neutrons collide with the nuclei of the atoms of the irradiated material, and there are basically three types of interactions (see §1.6, passage "Neutron radiation and its interactions") . For our purposes, the inelastic scattering of neutrons is important, in which the neutron transfers part of its kinetic energy to the nucleus and this causes an increase in its internal energy - excitation of the nucleus. When the nucleus returns to its original state (deexcitation of the excited nuclear levels), a gamma radiation photon is emitted with a precisely determined characteristic energy, given by the type of nucleus. These energies of secondary g- radiation from excited nuclei range from tens of keV to about 6 MeV. By spectrometric detection of this g- radiation it is possible to determine which elements are represented (according to the energy of the line g) and in what relative concentration (according to the intensity - the number of photons in the respective peak). The spatial distribution of these g-emitting nuclei can be determined by gammagraphic detection (gamma camera) *). Or with a spectrometric detector we can detect gamma radiation at different angles. Computer maps of spatial distribution of concentrations of specific chemical elements in the examined tissue can be created by computer reconstruction of positions (angles) and energies of this neutron-stimulated g- radiation.
*) The energy of g- photons emitted from excited levels of stable nuclei during neutron excitation is usually too high for imaging by standard gamma cameras. They are hundreds of keV to several MeV (eg for 16O the Eg = 6MeV, for 12C is Eg = 4.5MeV). For such energies, gamma camera collimators have poor spatial resolution and luminance, and the scintillation crystals used are too thin to achieve reasonable detection efficiency; also the spectrometric properties are not good. Therefore, spectrometric detectors not providing spatial information, but only the energy spectrum, are used in current experimental methods. Information on the position of the analyzed atoms is obtained either by rotating the detector equipped with a collimator and scanning from different angles, or by rotational scanning of the examined object by closely collimated neutron beams. In this second method, the path of the neutron beam defines the geometric position of the examined volumes and the spectrometric detector integrally scans all photons emitted by excited nuclei along the path of the neutron beam (ie the part of the photons that enters the detector). Unwanted background pulses can be significantly reduced by using a coincident spectrometer circuit, triggered by the pulse mode of a neutron generator. This is followed by computer reconstruction of data from individual projections. The resulting image is basically 4-D : for each voxel of the 3-D image, information about the energy of g- radiation ® representation of various elements is also stored. By selecting a specific energy (energy window), an image of the distribution of the corresponding specific element is obtained.
Note: For the gammagraphy of this hard
g- radiation, in principle can be used special Compton cameras (described above in the section "New and alternative physical and technical principles of gamm-ray imaging", passage " High Energy Gamma Cameras"), which are also still experimental...
  NSECT can in principle show the distribution of all elements and their isotopes, except hydrogen (whose nuclei do not have excited levels and therefore, there is no stimulated g- emission) and helium (which has too high an excitation energy of 25MeV). Neutrons are penetrating particles, so structures in the depth of the organism can be excited and displayed, with possible correction of the absorption (attenuation) of the primary neutron radiation as well as the registered stimulated g-radiation. The diagnostic potential of NSECT is due to the fact that the relative proportions of different elements, including trace elements, are different for different tissue types. It also differs slightly between healthy and tumor tissue. The method has so far been tested in the early diagnosis of breast and lung tumors. Apart from laboratory experiments, NSECT has not yet been implemented, it will probably remain only a physical-technical interest ...
 
Note: NSECT has some analogies and common aspects with other methods of neutron analysis of materials, especially neutron activation analysis NAA (INAA), described in §3.4, part "Neutron activation analysis". For special purposes of biological research is occasionally used neutron activation analysis in vivo: the relevant part of the organism is irradiated with neutrons (from a reactor or neutron generator), followed by a standard gammagraphic imaging of the distribution of induced beta radioactivity accompanied by gamma photons, mapping the distribution of the test substance in tissues and organs.


4.4. Gated phase scintigraphy
In our organism
(as well as in all higher animals) there are two important organs that work periodically in terms of time :
- The heart , which by regular contractions - systole and diastole of the ventricles and atria - acts as a "pump" to ensure the circulation of blood in the body.
- The lungs , which, through their breathing movements - shrinking and expanding, inhaling and exhaling - carry out the exchange of air and oxygenation of the blood in the body and the removal of carbon dioxide.
   These periodic events are not exactly regular and constant, their frequency fluctuates and varies individually, depending on the health and mental state and especially the physical load. The heart frequency at rest is about 55-75 pulses/min., the respiratory frequency around 15-20 breaths/min. However, with intense physical exertion, it can also increase 2-3 times - the need for faster blood pumping through the heart and faster exchange of oxygen and CO
2 by breathing.
   During this periodic activity, there are a relatively rapid movements of individual parts of the heart and lungs towards each other and with respect to the surrounding tissues and organs. These movements can be disturbing during scintigraphic diagnostics - they cause motion blur of the respective structures in the scintigraphic image.
   In scintigraphic analysis of own periodic actions in an organism, this periodicity can be advantageously used in a methodological approach called gated or triggers scintigraphy. In addition to scintigraphic impulses, another electrical signal is also recorded from the camera - an ECG or a respiratory signal - which suitably controls (triggers, gates) the course of the acquisition.

Phase scintigraphy of fast periodic processes - cardiac activity
Dynamic scintigraphy of the rapidly time-varying distribution of radioactivity encounters fundamental physical and technical problems. In order to faithfully capture the dynamics of the monitored process, it is necessary to use the best possible time resolution, ie a high frequency of short-term frames. The statistical character of radioactive decay then leads to significant statistical fluctuations of the measured pulses in the image. Relative statistical fluctuations are given by the expression 1/ÖN, where N is the number of pulses in the image cell accumulated over one frame. At high time resolution, the storage time of one image is very short (of the order of 10-2 s), the numbers of accumulated pulses are small and statistical fluctuations are very large *). The solution is not an enormous increase in applied radioactivity (this is usually not possible for other reasons, especially radiohygienic), because due to the dead time, the detection device is not enough to effectively process such a fast flow of pulses.
*) For accurate capture of cardiac activity, it is necessary to divide the heart cycle into very short time intervals; since the cycle lasts approximately 1 second, the acquisition time of one frame should be approximately 0.03 seconds. With this extremely short measuring time, the statistical fluctuations of the recorded pulse frequencies are so large (tens of %) that they do not allow the individual images to be evaluated. In such pictures, it would not even be possible to know which organ it is - only a shover of chaotically scattered dots would be visible.
   Thus, at first glance, it seems completely impossible to perform a detailed dynamic scintigraphy of one cardiac cycle. Fortunately, however, there are two favorable circumstances :
1. Cardiac activity is a periodic event (this is true at least approximately) ;
2. Cardiac activity is accompanied (or triggered) by electrical signals, that can be detected externally .

  In the case where the observed event is periodic, ie the distribution of radioactivity is a periodic function of time, the situation is significantly more favorable. If we denote the scintigraphic response function f(x, y, z, t), where x, y, z are positional coordinates, t is time, then the following will apply to the periodic process: f(x, y, z, t) @ f(x, y, z, t + k.T), k = 0,1,2, ...., T is the period ("@"means that equality is valid only on average, except for statistical fluctuations in decay and registration). The dynamic scintigram of such a process is then in principle given by the scintigraphic sequence of images of only one period (cycle). And conversely, the periodicity of the process offers the possibility to create a dynamic study of one cycle (periods) with a very high time resolution and at the same time with satisfactory "statistics": we measure several hundred individual cycles in succession with a high time resolution as they follow each other, and then synchronously compose (added up) the results frame by frame based on periodicity to create dynamic study of only one cycle :
         F
N (x, y ,z ,t)  =  k=1SN f [x ,y, z, t+(k-1).T] , t Î < 0,T )  .
Such a synchronously composed study F
N(x, y, z, t) will be called a phase dynamic scintigraphic study - it is a study of one "average" or "representative" cycle, composed of N common cycles of periodic process.
   This can be done directly (without additional information) using a computer if the period T is exactly known and constant. In practice, however, this is usually not met, eg the heart rate fluctuates somewhat. From statistically strongly scattered data, the computer does not completely "recognize" the individual phases of the periodic event and then has nothing to compose synchronously. Therefore, it is also necessary to input certain synchronization pulses (marks) to the computer from the outside, which make it possible to pinpoint the end of one cycle and the beginning of the next cycle. In the case of cardiac activity, such synchronizing or gating pulses can be signals from the ECG (R-waves), by means of which the computer always "recognizes" the end of one and the beginning of the next cycle.

Fig.4.4.1. Left: Schematic of creating a gated phase dynamic scintigraphy of the cardiac cycle based on scintigraphic data from the camera and synchronization derived from the R-wave of the ECG.
Right: Images from many different heart cycles are added synchronously to form a series of images capturing a single average cycle.

The principle of construction of the phase dynamic study of the cardiac cycle is shown in Fig.4.4.1. The computer program divides the time interval between two R-waves into short intervals of length Dt = T/(number of frames per cycle), where the number of frames per cycle is usually chosen 32, sometimes 16. Signal derived from the R-wave (electronically from its leading edge) determines the beginning of the cardiac cycle. Scintigraphic pulses corresponding to the beginning of the cardiac cycle in the interval 0 to Dt are recorded in the first image in the computer's memory. In the time from Dt to 2.Dt, the pulses are stored in the second frame, in the time from 2.Dt to 3.Dt to the third frame, etc. In the same way, the duration of the next cardiac cycle is divided, the beginning of which is signaled to the computer by another R-wave from the ECG; pulses registered during the first interval Dt are added to the first image of the previous heart cycle, pulses registered from Dt to 2.Dt are added to the second image of the previous cycle, etc. This process is repeated many times, forming a phase dynamic study of one average cardiac cycle in the computer's memory.

Cycles selection and excluding
As is well known, the heart rhythm is never completely regular, the heart rate and the period are more or less variable, even arrhythmias can occur. Only those cycles whose period does not differ too much from the average period need to be taken into account in the calculation. Irregular cardiac cycles, ie those whose duration is different from normal regular cycles, should be excluded from the record. These false cycles, caused by extrasystoles or other heart rhythm disorders, would distort the overall dynamics of the average cycle - it would no longer be a representative cycle. The limits for selecting the "correct" cycles are usually chosen to be
± 10% of T. It is necessary to exclude not only such an incorrect cycle with an anomalous period, but also cycle following it, as it may not begin at the correct stage of end-diastole. Due to the slightly fluctuating cycle length, the accumulated number of pulses in the last phase images is artificially reduced. In order not to distort the dynamics of the terminal section of the phase curves, an appropriate correction is made based on the number of cycles that contributed to the individual phase frames (the last few points of the phase curves are multiplied by factors > 1, inverse to the ratio of the number of cycles contributed to these last phase frames).

LIST-mode acquisition
The method of acquisition into image matrices described above is sometimes referred to as frame-mode. In earlier generations of acquisition computers, where a sufficiently large operating memory was not available, acquisition in the so-called LIST-mode was used: the coordinates (x, y) of individual pulses were sequentially stored in the memory as they came one after the other. Synchronization pulses from the ECG were also recorded into this continuous data stream under appropriate coding. Conversion to scintigraphic images (re-framing) and construction of the own phase study was then performed additionally. The advantage of LIST-mode was that the data could be subsequently re-framed with different selection of correct cycles, which is important in some heart rhythm disorders (such as bigimenia), when the correct phase study cannot be obtained in frame-mode. LIST-mode has been practically abandoned for dynamic scintigraphy for many years, but now it is being used for some iterative tomographic methods
(§4.3, section "Computer reconstruction of SPECT", "Reconstruction of PET images", "TOF - time localization of the annihilation site") .

First-pass phase scintigraphy
The above-described method of construction of phase dynamic scintigraphy of the cardiac cycle is performed in situations where blood carrying a radioindicator (eg
99mTc- labeled erythrocytes) is evenly and steadily mixed in the bloodstream - the so-called stady-state blood pool method. We will briefly mention the construction of phase scintigraphy of the cardiac cycle using the first-pass method, where the radioindicator is applied to the circulation as a compact bolus. Acquisition from the camera into the computer's memory starts when the bolus arrives in the heart chamber and ends before recirculation begins (when the structures would already overlap).

Fig.4.4.2.
Construction of a phase dynamic study of the cardiac cycle in first-pass radionuclide ventriculography. The "pulsated" curve represents the time course of radioactivity in the left ventricle during the first bolus flow.

Fig.4.4.2 shows the time course of radioactivity in the left ventricle during such a measurement. The curve is "pulsated" due to the periodic filling and emptying of the chamber by blood, bearing radioactive bolus. The construction of the phase dynamic study of the cardiac cycle is performed in a similar way as with the steady-state method, but only a few cycles can be taken into account - starting with the bolus arrival in the left ventricle and ending with the onset of recirculation, when the images of the chambers would already overlap.
  The advantage of the first-pass method is that the radioactivity is contained only in the chamber, so on the one hand the problem of correction on tissue and blood background is eliminated, on the other hand it allows scintigraphic "view" of the heart chamber even in such directions, in which at steady-state method would overlap the images of the right and left ventricles and possibly and other structures. The main disadvantage of the first-pass method compared to the steady-state method is the small number of cycles from which the phase study is created; therefore, the image quality is not very good due to statistical fluctuations. In addition, such a study may not represent a representative cardiac cycle, as selection of the "correct" cycles is not feasible here and acquisition is performed immediately after injection of a radiolabel, when central hemodynamics are affected by stress. From the quantitative parameters, therefore, only the ejection fraction can be objectively evaluated. The first-pass method is now rarely used to construct phase scintigraphy of the cardiac cycle. Used only bolus radiocardiography
(§4.9.4, section "Dynamic bolus angiocardiography").

  The method of phase (gated) scintigraphy is practically exclusively used in nuclear cardiology (§4.9.4 "Nuclear cardiology"). It is both radionuclide ventriculography (§4.9.4, the "equilibrium gated ventriculography"), whose comprehensive analysis program VENTR is described in §3.1 "Radionuclide ventriculography" books "OSTNUCLINE - Comprehensive assessment of scintigraphy", in recent years, the mainly SPECT myocardial perfusion (gated myocard SPECT - §4.9.4, part "Scintigraphy of myocardial perfusion").

Author's note :
We have been engaged in research and development of methods of scintigraphic phase dynamic studies with high time resolution at our workplace since 1976, practically in parallel with the development of these methods in leading laboratories in the world. An electronic device for R-wave detection and implantation of synchronization pulses into a computer was designed. We developed a program for the reconstruction and mathematical evaluation of radionuclide ventriculography on a small computer device Clincom
(operational memory only 12k!), which was probably the most complex procedure in this area at that time; then served as the basis for a comprehensive program VENTR ( "Radionuclide ventriculography ") on the GAMMA-11 device and later on a PC, the OSTNUCLINE system. Several dynamic phantoms were also constructed for this research and development work (starting with a rotating gramophone disc and ending with a flexible dynamic phantom of heart pulsation and circulatory pumping); some of them are described in the work "Phantoms and Phantom Measurements in Nuclear Medicine", part.4. "Dynamic phantoms".


4.5. Physical parameters of scintigraphy - imaging quality and phantom measurements
The task of scintigraphy is to provide a quality, ie objective, detailed and accurate imaging of the distribution of radioactivity in the examined object, both spatially and temporally (dynamic scintigraphy). We have mentioned above several limitations and adverse effects of a physical and technical nature that limit the possibilities and quality of scintigraphic imaging
("Adverse effects of scintigraphy"). To assess the quality of scintigraphic imaging, its optimization and detection of possible errors and defects, it is necessary to analyze and test the physical properties of scintillation cameras. As with any complex measuring instrument, the scintillation camera's properties can be described by several physical parameters :

Spatial resolution of a gamma camera
A scintillation camera is an imaging device, so the most important parameter is its distinguishing ability, or spatial or positional resolution (given in length units - millimeters) :

Spatial resolution
The spatial resolution of a scintigraphic image is called the smallest distance [mm] of two point radioactive sources in the displayed object, which are still distinguishable from each other in the scintigraphic image as two different objects .
Equivalent definition :
By spatial resolution we mean the width of the FWHM profile in the image of a point or line source in the middle of the maximum height of this profile, converted to the spatial scale in the object [mm] .

The spatial resolution is thus given by the half-width of the image profile of the point or line source; it is called FWHM (Full Width at Half Maximum). Two point radiation sources can be distinguished from each other in the scintigraphic image, only if the distance between them is at least FWHM - Fig.4.5.1 :


Fig.4.5.1 Spatial resolution of scintigraphic imaging - analysis of images of point sources, which are displayed as "blurred" scattering circles.
a) By a "blurred" image of a point source we lead a profile curve PSF (Point Spread Function), its half-width FWHM indicates the resolution of the display. b, c, d, e) Images of two point sources located at different distances from each other. If this distance is greater than FWHM, these sources are displayed as two separate objects. When approaching in to a distance equal to FWHM and smaller, this two images already merge into one - the sources can no longer be distinguished from each other. g) Schematic representation of the interweaving, superposition and merging of images of two closely spaced point sources (on PSF profile curves) - corresponds to the situation according to Fig. e) .

Note: It was measured with point sources with a diameter of approx. 1mm, 50MBq
99mTc, on a Nucline MB9201 camera close to the front of the HR collimator, matrix 256 x 256, zoom 2x, with strongly enlarged image sections, approx. 4x. Slight deformations of the circular shape of the images of point sources were caused by small geometric irregularities of the holes of the HR collimator (which are visible at this high magnification, but do not manifest themselves in practical scintigraphy).

In addition to FWHM, the resolution of the camera is sometimes characterized also by the parameter FWTM (Full Width at Ten Maximum), which is the width of the image profile of the line source in tenths of the maximum height of the profile. Of course, this value is higher: in situations without a scattering environment, it is approximately FWTM @ (1.8 - 2) x FWHM, in the presence of a tissue (or aqueous) scattering environment, it is approximately FWTM @ (2 - 2.5) x FWHM.
Spatial resolution FWHM <--> modulation transfer function MTF 
It should be noted that the FWHM value determined with PSF or LSF may not to capture the quality of the display completely objectively in terms of resolution! Scattered radiation and cross-radiation trought the collimator septa mainly expands the lower part of the PSF, lower than 50%
(see Fig.4.3.3d, and Fig.4.5.4e), so it may not affect the value of FWHM. For a completely detailed physical analysis of the distinguishing properties of imaging systems, the so-called modulation transfer functions (MTF) are used - see the work "Theory of scintigraphic imaging and modulation transfer functions", which take into account all points of PSF.
Terminological note:
The spatial resolution of a gamma camera is sometimes abbreviated to resolution. Please do not confuse with spectrometry energy resolution (....) or time resolution (as the detector dead time is sometimes called "
Detector dead time") - it's something completely different..!.
   The spatial resolution of a classic single-photon gamma camera - planar, SPECT - is determined by two components :
1. Geometric resolution of collimator Rcolim ,
which depends on the hole diameter d and the length of the holes - channels L of the height (thickness) of the collimator, significantly also on the distance h of the displayed object from the collimator :
                        R
colim ~  d. [1 + (h + m) / L]  ,                       (4.5.1)
where the width of the gap m between the rear face of the collimator and the camera crystal
(its center) is still manifested - Fig.4.5.2a. The resolution of the collimator is manifested by the fact that each point (point source) in the object is geometrically projected on the camera detector as a small blurred circular trace with a diameter dependent on the Rcollim - Fig.4.5.2b.

Fig.4.5.2 Two basic components of the spatial resolution of a gamma camera - a collimator and a crystal + photomultipliers.
a): Trigonometric analysis of the collimator resolution. From each hole of the parallel collimator we can draw an imaginary cone defining the area from which gamma radiation can pass through this hole to the camera detector
(radiation from places outside this cone is absorbed by the lead septa of the collimator). As the distance from the collimator increases, this detection cone widens, thus deteriorating geometric spatial resolution of the image projected by the collimator onto the scintillation crystal of the gamma camera.
b): The resolution of the collimator is manifested by the fact that each point (point source) in the object is geometrically projected on the camera detector as a small blurred circular trace with a diameter depending on the distance from the collimator.
c): Approximate representation of the internal resolution of the camera detector, related to statistical fluctuations of detected scintillations, blurring X, Y coordinate pulses and Compton scattering of gamma photons.
d):
99mTc point source display at different distances h from the front of the collimator HR - degradation of FWHM resolution with distance ..

The spatial resolution of the collimator improves with increasing the length of the holes - channels (thickness of the collimators) and decreasing the diameter of the holes. Collimators with small and longer holes have better resolution (and at the same time less dependence on distance). The spatial resolution of the gamma camera significantly deteriorates with the distance *) of the displayed structure from the front of the collimator (Fig.4.5.2b, and Fig.4.5.6 on the left). The gamma camera (front of the collimator) should therefore be placed as close as possible to the surface of the patient's body.
*) For collimators with a different arrangement of holes (described above in §4.2, section "Scintigraphic collimators") the geometric situation is more complicated, but basically the same rule applies to the deterioration of the spatial resolution at a greater distance from the collimator face - Fig.4.5.6 on the left.
Transverse radiation trough the collimator partitions 
For the correct imaging function of the collimator, it is necessary that gamma radiation passes only through the holes, while the partitions between them would not transmit radiation. However, this cannot be fulfilled 100%, a certain part of the radiation penetrates even through the septa. Especially at higher energies gamma may cause transverse radiation of gamma rays trough collimator septa, if these partitions between holes are thinner than is optimal for the gamma energy used. Transverse radiation trough the septa impairs the contrast of the scintigraphic image. If the low-energy LEHR collimator, optimized for 140keV
99mTc, were used for scintigraphy with 111In or even with 131I., we would get an image with greatly degraded contrast! The effect of transverse-radiation trough of septa and its influence on the quality of scintigraphic imaging is analyzed in Fig.4.5.3 :


Fig.4.5.3 Transverse radiation of gamma photons trough septa of a collimator and its influence on the quality of a scintigraphic image.
a) Schematic representation of the aborption of low-energy gamma radiation in lead septa between the collimator orifices and the partial transverse radiation of high-energy gamma through the collimator septa.
b) The transverse radiation of photons of higher energy trough the collimator septa causes the expansion of the edge parts of the point source image
(red part of the PSF curve).
c)
, d) , e) Images of point sources of radionuclides emitting various energies of gamma radiation -
99mTc (140keV), 111In (245keV), 131I (364keV), recorded with a LE HR collimator for low energies.
f)
Phantom Jasczak, filled with
111In and displayed with a gamma camera with a collimator Medium Energy (top) and a low-energy collimator LE HR (bottom). Radiation trough the LEHR septum completely degrades the contrast of the image here - an almost invaluable scintigram.
Note: The typical "star-shaped artifact" of diagonal transverse radiation is due to the geometric configuration of the holes and lead partitions, which are projected somewhat narrower in the diagonal direction. For different types of collimators, the shape of this artifact may be somewhat different, depending on the details of the geometric design of the openings and partitions.

The proportion of radiation penetrating the barrier between the apertures - transmission factor, is e-m .stran, where m is the linear absorption coefficient of the collimator material (lead) for the required gamma energy and stran is the shortest path that gamma radiation can penetrate the barrier from one hole to adjacent hole. For a collimator with the diameter of the holes d , their length L and the thickness of the baffles s, the shortest path of transverse radiation trough the baffle stran = s.L/(2d + s), so that the transmission factor is e-m .s.L/(2d + s). Optimization of partition thicknesses between the holes is performed on the basis of the requirement for a sufficiently low value of the transmission factor (most often 0.05); this leads to a condition for the transmission factor e-m .s.L /(2d + s) < 0.05 (described in §4.2, passage "Scintigraphic collimators"). This gives a limitation for the thickness of the collimator baffles s > (6.d/m) /[L - (3/m)].
Effective length of the collimator holes
 
In the formula (4.5.1) for spatial resolution, we did not consider transverse radiation trough the collimator septa. If transverse radiation occurs, it causes a seemingly effective shortening of the actual length of the collimator holes in terms of geometric collimation. Therefore, the so-called effective length of collimator holes L
ef = L-2.m-1 sometimes introduce, where m is the linear coefficient of attenuation in the collimator material (lead); it is a correction for the transmission of photons through two opposite baffles between the holes. The introduction of Lef instead of L in (4.5.1) describes the deterioration of resolution by transverse radiation. However, this is not very important, because the PSF does not have a Gaussian shape here, it is widened in the lower part and leads to a deterioration of not so much the FWHM resolution value, as the image contrast, as seen in Fig.4.3.5d, e).
2.
Internal resolution of the camera detector Rint
Internal (proper, intrinsic) resolution is given by the accuracy with which the system of photomultipliers and related electronics is able to locate the position of scintillation in the crystal. At the quantum level, the internal resolution is influenced by statistical fluctuations in the production of light photons after the interaction of g- radiation in the detector and variations in the number of electrons emitted from the photocathode and dynodes of the photomultipliers. These fluctuations "blur" the amplitudes of electrical signals from photomultipliers and thus the values of the resulting coordinate pulses X, Y - Fig.4.5.2 c. Using a larger number of photomultipliers with higher quantum efficiency and good optical contact with the crystal leads to somewhat better internal resolution.
   The thickness of the crystal also has a negative effect here, in two ways. On the one hand, it is a geometric "blur" of the positions of light flashes registered by photomultipliers - scintillations occur at different depths, and the system of photomultipliers evaluates their positions X, Y somewhat differently. Furthermore, it is the multiple Compton scattering of
g- photons in the detector, which also causes uncertainties in the X,Y-localization of the interaction site of primary gamma-photons. A thinner crystal and a larger number of photomultipliers allow you to achieve better resolution. For these reasons, thin scintillation crystals about 0.7-1.8 cm thick are used. Current scintillation cameras, optimized for g 140keV, have a crystal thickness of mostly 9.5mm and achieve an internal resolution of 2.5-3.5 mm.
The total resolution of the R gamma camera (external - extrinsic) 
is then given by the geometric sum of both subcomponents: R = Ö(R2int + R2collim). However, in practice, the total resolution is not calculated according to this relationship, but is measured using the point or line sources placed in the field of view of the camera with the given collimator ("Phantoms and phantom measurements in nuclear medicine", section "Measuring the positional resolution of the camera").
   In practice, it is given in the first place for each camera its internal resolution, which is determined from the production technology - typical values of internal resolution for newer cameras are about 2.5-4 mm. The internal resolution represents the limit value of the resolution, below which at the given camera can no longer be reached, and which can only be approached using an ultra-high resolution collimator.
   Furthermore, the total resolution of the camera with individual collimators for the given distances of the source from the collimator (usually 10 cm), or even in the presence of a scattering environment
(scattering environment between the radiation source and the detector - water and tissues, which is always present in scintigraphic imaging of radioactivity in the body, somewhat worsens spatial resolution and more markedly worsens a contrast in the image).
   These values of the total resolution are different, for LE HR collimators (high resolution), optimized for 140keV
99mTc, at a distance of 10 cm from the front of the collimator, it is about 8-10 millimeters. For collimators for higher energies, the overall resolution (in 10 cm) is worse, about 12-15 mm. As analyzed above, the spatial resolution of the gamma camera significantly deteriorates with the distance of the displayed structure from the collimator front. The gamma camera (front of the collimator) should therefore be placed as close as possible to the patient's body surface when scanning.
Influence of scattered radiation 
Compton scattering of gamma radiation in the material environment - in the patient's tissue, also contributes to the deterioration of the image quality (the physical nature is described in §1.6, section "Interaction of gamma radiation and X", passage "Compton scattering"). Scattered radiation primarily impairs the contrast of scintigraphic imaging, to a lesser extent the spatial resolution of FWHM. This effect is analyzed in Fig.4.5.4 :


Fig.4.5.4 Compton-scattered gamma radiation in scintigraphy and its influence on the quality of scintigraphic image.
a)
If, coincidentally, the photon is scattered in the tissue at such an angle that the scattered photon passes through the collimator orifice and is detected by the camera crystal, then these scattered photons g´ can be detected from a false location - is detected gamma photons seemingly coming from another places, from where it was originally radiated during the radioactive transformation.
b)
These false scattered photons
g´ have a lower energy than the "true" direct and primarily detected photons g (part of the energy was transferred to the electron e- when scattered in the substance), so they usually do not fall into the photopeak. By carefully setting the analyzer window to the photopeak of the given radiation g, we can therefore largely eliminate the Compton-scattered radiation g´.
c) , d), e) Images of point source without scattering medium (c) and with scattering medium in photopeak measurement (d), and including Compton scattering in a wide analyzer window (e).
f), g)
An image of the Jasczak water phantom with a narrow photopeak window and a wide window, including scattered radiation
(deterioration of the image contrast can be seen) .

Note: Spatial resolution of PET
For two-photon cameras of coincidence positron emission tomography PET, the physical principle of imaging and analysis of spatial resolution is different - it was described above in §4.3, passage "
Spatial resolution of PET" (spatial resolution of PET cameras is generally somewhat better than conventional Anger cameras).
The measurement of the spatial resolution of the camera 
can be performed in two ways :
¨ Quantitative physical measurement
- by analyzing images of point and line sources, through which we conduct profiles - sections, thus obtaining PSF or LSF curves. From them, we then directly determine the value of the half-width-resolution FWHM in [mm]
(or by a more complex analysis of the so-called modulation transfer function MTF .... "...") . For practical determination of the resolution, it is more appropriate to use a line source, by image of which we can conduct independently of several LSF profiles, or sum these profiles and thus achieve smaller statistical fluctuations.

Fig.4.5.5. Measurement of positional resolution of a gamma camera.
Left: Line source (capillary) and two point sources for measuring positional resolution.
Middle, right: The measurement was performed with a filling solution of 99mTc at distances of 0, 5, 10, 15 and 20 centimeters from the front of the collimator HR of the camera Nucline TH. The degradation of FWHM resolution with distance can be seen in the images and profile curves.

¨ Visual evaluation of phantom images
- most often they are so-called Bar-phantoms placed to a planar homogeneous source (mostly
57Co), or vessels of complex structure, filled with a 99mTc solution (eg Jasczak phantom ). It is a more or less qualitative evaluation, the value of the resolution is rather estimated from the images; however, it is usually sufficient for practice and comparison. Performing of both methods of resolution measurement is described in the work "Phantoms and phantom measurements in nuclear medicine", section "Measurement of positional resolution of the camera".

Sensitivity (detection efficiency) of a scintillation camera
Detection efficiency or sensitivity S of devices for detection and spectrometry of ionizing radiation - radiometers - is generally defined as the ratio between the number of detected pulses
(quanta registered by the detector) and the number of incoming radiation quanta; the relative and absolute efficiency is introduced, often expressed in % (physically defined and discussed in §2.1, section "General physical influences and instrumentation for the detection and spectrometry", passage "Detection efficiency and sensitivity").
   Detection efficiency of the classic single photon gamma camera - planar, SPECT - is determined by two components :
1. Geometric transmittance (aperture) of the collimator for gamma radiation
indicates, what part of the incoming gamma photons the collimator passes on to the camera detector
(unlike photons which are absorbed in the collimator septa). It depends in principle, on what part of the total area of the collimator is occupied by the through holes and what part of the absorbend lead baffles between them. According to the trigonometric analysis in Fig.4.5.2a, the geometric transmitance Scollim of the parallel collimator is :
        S
collim =  (d/L)2 . [d2/(d + s)2] . Kg . [100%]   ,                 (4.5.2)
where d is the diameter of the holes, s thickness of the baffles, L is the length of the holes-channels given by the thickness of the collimator. The geometric factor K
g depends on the shape and arrangement of the holes and partitions (for circular holes in a hexagonal arrangement is Kg = 0.24, for hexagonal holes in a hexagonal arrangement is Kg = 0.26, for square holes in a quangangular arrangement is Kg = 0.28). Conventional LE HR collimators have a geometric transmitance of about 1.2%, LEAP about 2%, HS about 2.2%. ..............
Note: In the equations for spatial resolution (4.5.1) and detection efficiency of the camera (4.5.2) we did not consider transilumination gtrough collimator septa. If it occurs, it leads to a deterioration of the image quality (see Fig.4.5.3), but at the same time to an increase in detection efficiency. However, this is a negative phenomenon, the quantification of which in physical parameters is irrelevant ...
Independence of gamma camera detection efficiency on distance
 
The intensity of the radiation decreases with the square of the distance from the source
(this is exactly the case for a point emitter). Therefore, the detection efficiency of conventional radiometers is significantly reduced for greater distances from the measured sources. However, for gamma cameras with parallel collimators, the detection efficiency (sensitivity) does not depend on the distance of the displayed source from the collimator front! In formula (4.5.1) the parameter h of distance does not appear. The display of the point source in a wide range of distances 0-30cm from the front of the collimator in Fig.4.5.2 d, although it does shows a deterioration of spatial resolution and decreasing image brightness, but the total number of pulses is the same in all images, area (integral) under PSF is the same for all distances.
  This surprising behavior is due to the specific properties of geometric collimation for parallel collimators. We can clearly illustrate this according to the schematic drawing in Fig.4.5.2 a,b as follows: As the source moves away from the collimator front, the number of photons incident on the individual holes decreases quadratically as 1/h
2. However, the number of holes through which radiation can pass to the detector increases quadratically in proportion to h2. These two opposing trends cancel each other out, so that the total photon flux - the efficiency of the collimator - does not change with the distance between the source and the collimator.
Note: This rule does not apply to special convergent or Pinhole collimators, the detection efficiency here changes significantly with distance - it increases or decreases
(see the section "Imaging properties of special collimators" below).
2. Internal detection efficiency of the crystal and photomultipliers of the camera Sint ,
indicating which part of the gamma photons incident on the detector
(i.e. passed through a collimator), is actually detected by the system of scintillation crystal, photomultipliers and analyzer, in the form of pulses creating the scintigraphic image. It depends on the thickness and conversion efficiency of the scintillator, the gamma radiation energy, setting the window width of the analyzer to the photopeak. The photopeak detection efficiency of a standard gamma camera detector with a 9.5 mm thick crystal for 140keV 99mTc is about 80%, for 364keV 131I then about 30%.
Overall - system detection efficiency 
- sensitivity of the camera S is then given by the product of both of these components S
collim . Sint ; while doing so the main determining component is the efficiency of the collimator. However, in gamma cameras, where the source of gamma radiation is a radionuclide, the sensitivity - detection efficiency - is usually quantified in a special way: as the number of pulses detected by the camera per unit time [per second] - cps, referred per unit of activity [kBq, MBq] of the radionuclide used in the displayed source; for the planar/SPECT scintigraphy is usually 99mTc, for the PET 18F. Only exceptionally is it expressed here in %.
   In scintigraphic diagnostics, we are mostly concerned with the relative assessment of the distribution of the radioindicator in various parts of the examined object. In the case of so-called quantitative scintigraphy, however, we may also be interested in the absolute activity of the radioindicator in the investigated area. In order to determine this real activity from a scintigraphic image, we need to know the efficiency (sensitivity) of radiation detection
g of the used radionuclide by a scintillation camera. We also need to know the detection sensitivity of the gamma camera to determine the optimal applied activity of the radio indicator to obtain sufficiently high-quality scintigraphic images.
   In the case of scintillation cameras, for practical use, the detection sensitivity is related to the radioactivity of the examined object :

Detection efficiency (sensitivity) of a scintillation camera
The detection efficiency, or sensitivity S, of the gamagrapfic system is quantified as the pulse frequency N[imp./s] measured by a scintillation camera with a point radiation source g (located at the desired field of view), relative to the activity unit A[MBq] of the source: S = N / A .
It is expressed in units [imp. s
-1 MBq-1 ], or [cps/MBq] or [cps/kBq] .

It is given for a specific type of radionuclide and collimator. Most often, the sensitivity for planar and SPECT cameras given for radionuclide 99mTc, for PET cameras 18F. For different radionuclides, the sensitivity of a scintillation camera generally has different values, depending on the yield of gamma photons [%] (number of gamma quanta/100 conversions of the radionuclide) and their energy [keV].
  
The basic physical measurement of the detection efficiency of a gamma camera - at a given distance and location of the field of view - is performed with a point source of the required radionuclide. We can thus perform detailed measurements of the dependence of the detection efficiency on the distance in different places of the field of view (as can be seen in the right part of Fig.4.5.6). Another way of measuring the "averaged" sensitivity is with a planar source, which lies entirely in the field of view (this method is especially suitable for cameras equipped with a parallel collimator) - in the image of the planar source are averaged event. local sensitivity inhomogeneities and measurement error may be somewhat reduced. According to the recommended NEMA procedure (to ensure good accuracy and reproducibility) we determine the sensitivity of the camera by placing a 10cm diameter bowl with 99mTc solution with exactly known activity in the middle of the field of view - approx. 10MBq, a layer of solution up to 1cm. We accumulate a scintigraphic image with the given collimator and in the required configuration (acquisition time min. 100sec), in which we determine the number of pulses in the ROI of the bowl image and convert it to 1MBq and 1sec.
   In addition to the actual detection efficiency of the scintillation crystal of the camera on the given gamma radiation, the resulting, total - system sensitivity depends decisively on the collimator used. For universal collimators of the LEAP type, the sensitivity of scintillation cameras for
99mTc is around 150-300 cps/MBq, for high-resolution (HR) collimators only about 50-100 cps/MBq 99mTc. As discussed above, for parallel collimators, the registered number of pulses is virtually independent of distance, while for convergent and Pinhole collimators. (§4.2, part "Scintigraphic collimators") the dependence of the detection sensitivity on the distance from the collimator face is very significant - Fig.4.5.6 on the right (it is derived below in the section "Imaging properties of special collimators"; the independence of the detection efficiency over distances is maintained with these special collimators only for sources with a homogeneous area distribution of activity exceeding the field of view).
   For PET positron emission tomography cameras, which use electronic coincidence collimation instead of mechanical collimators, the detection sensitivity is significantly higher, approx. 7000-10000 cps/MBq
18F - see §4.3, passage "Detection efficiency (sensitivity) of PET".


Fig.4.5.6. Dependences of the spatial resolution FWHM (left) and the detection efficiency of the S (right) gamma camera on the distance of the source from the front of different types of collimators. In the box on the far right, the entire detection efficiency curve for the convergent collimator is plotted up to a distance of 70 cm, capturing a significant maximum in the focus and the subsequent decrease.
Note: These curves are only approximate and are more or less illustrative. They were created by comparing and interpolating a series of measurements of point and line sources 99mTc with different collimators on cameras PhoGammaHP, Nucline MB9201 and TH, Symbia T. Specific values of resolution and sensitivity may differ slightly for individual cameras of different types and manufacturers, but dependency trends are captured objectively .

Influence of the material environment
The analysis of the detection efficiency (sensitivity) of the gamma imaging was performed from above in a situation without a substance-absorbing environment - in vacuum or in air. However, in practical scintigraphy, there is a tissue environment between the imaged structures with distributed radioactivity in the organism and the gamma camera, with which gamma radiation interacts, which mainly leads to the absorption and attenuation of gamma radiation. During the passage of the tissues from the point of origin towards the camera detector, a certain amount of radiation
g is absorbed during the interaction with the tissue substance - due to the photoeffect and the Compton scattering in the tissue. It leads to exponential decrease in the count frequency of N detected photons g with increasing depth h of the radiolabel distribution in the body: N = No .e -m .h, where m is the linear attenuation coefficient, depending on the radiation energy g and on the tissue density (for g 140keV 99mTc this absorption coefficient is m @ 0.15 cm-1). This loss of gamma-ray by absorption, also called attenuation, is manifested in scintigraphic images by an artificial reduction the number of pulses from structures deposited at greater depths, compared to structures closer to the surface. In such a case, the statement that the detection efficiency (sensitivity) does not depend on the distance of the displayed source from the (paralell) collimator face no longer applies. Here, the detection efficiency decreases significantly with the distance - depth - of the displayed source !
   An approximate dependence of S ~ R-2 applies between the sensitivity of the camera S and its total spatial resolution R (= FWHM) (for parallel collimators it follows from the comparison of formulas (4.5.1) and (4.5.2) ). So the better the resolution of the imaging system, ie the smaller R = FWHM, the lower its sensitivity - and vice versa. When trying for high resolution (using a UHR collimator), this leads to a lower pulse density in the image and therefore to higher statistical fluctuations (higher noise). In other words, resolution and sensitivity they compete with each other - collimators with better resolution have lower sensitivity and vice versa.
Imaging properties of special collimators - convergent and Pinhole 
Imaging properties - display scale, spatial resolution and detection efficiency (as well as linearity) - of these special collimators differ significantly from basic collimators with parallel holes. Above all, collimators with parallel holes projected the displayed radioactive structures onto the detector crystal in an unchanged size - on a 1:1 scale, while special collimators provide an enlarged or reduced display, depending on the distance of the source from the collimator: just this dependence of the magnification ratio - "optical zoom" is the main reason for their use. Spatial resolution it although does basically worsens with distance, but at a different "pace" than with parallel collimators. The detection efficiency (sensitivity) of parallel collimators is practically independent of distance, while that of special collimators increases or decreases significantly with distance
(as can be seen in Fig .4.5.6 on the right). We will give an overview of the imaging properties of these collimators according to Fig.4.5.7 :


Fig.4.5.7 Geometric arrangement of holes and imaging properties of collimators: Parallel, Convergent and Pinhole.

Parallel collimator
Display scale: M = 1 - projects an image without resizing on the camera detector, display 1 : 1 .
Spatial resolution
: R
colim ~ d. (1 + h/L) (formula (4.5.1) ) - deteriorates with increasing distance from the collimator, approximately linearly.
Detection efficiency (sensitivity)
: S
collim = (d / L)2 . [d2 / (d + s)2 ]. Kg . [100%]  (formula (4.5.1)) is independent of the distance from the collimator. The imaging properties of parallel collimators were discussed in detail above at the beginning of this chapter "Physical parameters of scintigraphy" in the sections "Gamma camera resolution" and "Gamma camera detection efficiency".
Convergent collimator 
Display scale:
M = (f + L)
/(f.h) - provides magnification of the image depending on the distance h of the source from the collimator.
Spatial resolution
: R
colim ~ [d . (L + h) / L] . (cos q)- 1 . [1 - (L / 2) / (f + L)] deteriorates with distance in a manner similar to a parallel collimator.
Detection efficiency (sensitivity) : S
collim = (d / L)2 . [d2 / (d + s)2 ] . [f2 / (f -h)2] . Kg . [100%] . With a convergent collimator, the detection efficiency increases with distance from the collimator. Significant maximum - up to 30 times higher than at the front of the collimator! - reaches in the focus (which is usually at a distance of about 40-60 cm), from where gamma radiation passes into the detector through all the holes of the collimator; behind the focus, the detection sensitivity decreases - see the curve in the box in Fig.4.5.6 on the right. However, this area of long distances is not usable for practical scintigraphy, as there is already poor spatial resolution (and the structure of holes and partitions can be disturbing, when is magnified projected on the detector) .
   Convergent collimators show an optimal combination of resolution, efficiency and display scale at a distance of about 15-20 cm, which corresponds well to the size and depth of the heart. They were therefore often used in nuclear cardiology - at ventriculography and scintigraphy of myocardial perfusion
(§4.9.4 "Nuclear cardiology").
   
Note: The opposite - divergent - hole configuration is in divergent collimators (now no longer used), which provide image reduction. A divergent collimator basically arises when we turn the convergent collimator and "deploy it in the opposite direction" on the camera detector.
Pinhole collimator
 
Display scale: M = L / h - provides reduction or enlargement of the image, according to the distance h. The image is mirror- inverted. For larger distances h > L the image is reduced, for smaller distances h < L from the collimator the image is enlarged.
Spatial resolution: R
collim = d . (L + h) /L is very good at small distances (with a hole diameter d = 2mm, at distances up to 10cm, a resolution of Rcollim approx. 3mm and a total FHHM resolution of around 4-5mm is achieved). At greater distances, the resolution deteriorates, but is still better than other collimators.
Detection efficiency (sensitivity) : S
collim = (d / 16.h2 ). cos3 q . The detection sensitivity of a gamma camera with a Pinhole collimator is relatively high only in close proximity to the aperture (at 3 cm it is about 500 cps/MBq), but it decreases sharply with distance. At greater distances it is already very low (at 10 cm it is about 50 cps/MBq, at 20 cm only 10 cps/MBq), for practical scintigraphy it is completely insufficient (perhaps only for monitoring high therapeutic activities 131I) .
   Due to these imaging properties, the Pinhole collimator is suitable for scintigraphy of small organs, especially the thyroid gland - see
§4.9.1 "Thyrological radioisotope diagnostics", where it provides an enlarged image with very good resolution and sufficient detection efficiency.
   Convergent and Pinhole collimators show a certain inhomogeneity and nonlinearity of the image- resolution and detection efficiency differ somewhat at different points of view, even at the same distance h from the collimator front. These differences depend on the radial distance of the display source from the axis of the collimator, which is expressed at a trigonometric analysis of the angle
q between the collimator axis and the line connecting the source and its image. Deviations in the values of resolution and sensitivity are given by the cosine of the angle q. In practical scintigraphy, while maintaining optimal configurations, these deviations are not very significant...
   Imaging properties of the collimators with different geometric arrangement of the holes are most clearly seen at images of linear orthogonal grid (its construction is described in "Phantoms and phantom measurements in nuclear medicine" image "Grid") :

For a collimator with parallel holes (such as LE HR left) we get a linear imaging of the grid everywhwre, only for a greater distance from the front of the collimator, the spatial resolution deteriorates (blured grid). With a convergent collimator (such as a SmartZoom with the convergent center part) the image of the center part increases with increasing distance. With the Fan Beam collimator (which is convergent in the transverse direction, parallel in the axial direction), the grating espands only in the transverse direction with increasing distance, it remains the same in the axial direction.
   The most striking dependence on the object distance exhibits the collimator Pinhole: tightly close to the opening we get the image magnified many times, with increasing distance the zoom decreases and for distances above approx. 20cm the image is already reduced.
   Of all the images is also seen a general trend of deteriorating resolution (and thus contrast in the image) with the distance from the front of the collimator.

Homogeneity (uniformity) of the camera's field of view
The scintigraphic image is created in the gamma camera in a very complex way, the signals go through a number of precisely tuned electronic and electro-optical components. However, the individual links in this chain may show some changes, which may cause deviations and defects in the images created.
   Another important parameter of the quality of scintigraphic imaging - homogeneity (also called uniformity) indicates, whether individual places in the field of view are imaged with the same efficiency (sensitivity). By inhomogeneity of the displayed field we mean local artificial changes in accumulated number of pulses, caused by local changes in the sensitivity and linearity of the image. Inhomogeneity is usually caused by different sensitivity of individual photomultipliers or their different spectrometric settings, deviations in the adjustment of electronic circuits, defects or inhomogeneities in the collimator or scintillation crystal.
   Field homogeneity characterizes the camera's ability to provide an accurate
(ie, homogeneous) picture of a homogeneous distribution of radioactivity. By irradiating the camera's field of view with a homogeneous flux of photons of radiation g, we obtain an image of a homogeneous source, which should also be completely homogeneous (except for statistical fluctuations). Possible inhomogeneities in this image are seen visually, but they can also be expressed quantitatively, mostly in percentages :

Homogeneity of the camera's field of view (integral)
The homogeneity of the camera's field of view is the maximum deviation of the actual image created in response to the homogeneous irradiation of the camera detector, from the ideally homogeneous image :
H = 100
[%] . (N
max - Nmin ) / Nmean ,
where N
max is the maximum, Nmin minimum and Nmean the average (mean) number of pulses accumulated in the pixels of the homogeneous source image.

The overall homogeneity of the field of view thus defined is referred to as integral homogeneity. Since the human eye is more sensitive to differences in the brightness of neighboring areas in the visual assessment of images, the so-called differential homogeneity may also be useful for evaluating the homogeneity of the image. The following criterion was adopted for its quantification :
   Differential homogeneity is the ratio of the largest difference in the number of pulses in adjacent cells (row and column) in the homogeneous source image, divided by the average number of pulses in the image H
dif = max(Ni - Ni-1) / Nmean . To reduce the effect of statistical fluctuations, the determined number of pulses is averaged over 5 cells.
Whole and central field of view 
It follows from the design of the scintillation camera, that the quality of the scintigraphic image is usually best in the central part of the field of view, while in the peripheral parts it may be somewhat degraded. Therefore, homogeneity
(and sometimes other camera parameters) is often determined separately for the entire field of view (UFOV - useful field of view) and separately for the central part of the field of view (so-called CFOV - central field of view). 75% of the entire field of view is usually taken as the central part. For quality and correctly adjusted (calibrated, tuned) cameras, the integral homogeneity in the central field should not be worse than about 3,5%, and in the whole field of view up to 5%; differential inhomogeneity in the central field should be in the range of 1.5 - 3%.
   Similar to resolution, the homogeneity of the scintillation camera is recognized by :

¨
The internal homogeneity of the camera detector (intrinsic) 
- is given by the homogeneity of the scintillation crystal and its light response, light collection, sensitivity and adjustment of individual photomultipliers. It is measured by homogeneous irradiation of a crystal without a collimator.

¨
Overall homogeneity of the camera ("external" - extrinsic) 
- given the internal homogeneity of the camera detector
+ homogeneity (or inhomogeneity, defects) of the used collimator. It is measured with a collimator attached, a homogeneous flat source is displayed, most often 57Co.
Visual field inhomogeneity correction 
The inhomogeneity of the field of view of the camera can be reduced or eliminated in two steps :
1.
Careful adjustment - matching the individual photomultipliers to the same detection efficiency (same photopeak position), so-called tuning. In earlier analog cameras, this was done manually using potentiometers in the preamplifiers of the individual photomultipliers, with check on the oscilloscope screen. Current digital cameras have a computer procedure, which for each photomultiplier uses ADC <--> DAC converters to adjust the gain so that the top of the photopeak of the radionuclide is exactly in the middle of the set analyzer window.
2. Computer correction using a suitable matrix of correction coefficients g ij to correct the remaining inhomogeneous response of the detector - Fig.4.5.8. The accumulated numbers of pulses A ij in the individual elements (i, j) of the original uncorrected image are multiplied by the correction coefficients g ij from the correction matrix, thus creating an image *A ij corrected for inhomogeneity :
            *A
ij   =  A ij . g ij   .

Fig.4.5.8. Computer correction of gamma camera image inhomogeneity.
Left: Image a ij of a homogeneous source, showing significant inhomogeneities.
Middle: Matrix of correction coefficients g ij .
Right: Multiplying by correction factors creates a corrected image *a ij that is already homogeneous.
Note: Instead of the usual luminance modulation, an isometric display is used here, where the height of the elements (pixels) above the base is proportional to the number of pulses contained. The curves at the top are cross-sections, taken through the center of the images.

The matrix of correction coefficients g ij is obtained from the scanned image h ij of a homogeneous source (which we know should ideally be homogeneous - constant) as its normalized inverse matrix. The correction coefficients g ij for individual elements (image cells - pixels) i, j are calculated as the ratio of the average number [pulses/pixel] in the whole field of view W to the number hij [pulses/pixel] in a given image location (i, j) of a homogeneous source :
            g
i j   =   ( i, j Î W S h ij ) / (N. h ij )  ,
where N =
WS i +WS j is the total number of elements of the visual field image W. In places (i, j) of the field of view with reduced detection efficiency, the correction coefficients g ij are slightly higher than "1", in places of higher efficiency they are slightly lower than "1".
   The image of the homogeneous source must be obtained under the same physical conditions *) as the images that we want to correct with the resulting matrix. This correction matrix g
ij is stored in the memory of the acquisition computer and its accumulated number then multiplies the accumulated number of pulses at a given location of the field of view during scanning (acquisition).
*)
On what the homogeneity of the field of view depends :
The homogeneity of the gamma camera imaging depends (in addition to the mechanical, detection, optical and electronic properties of the device) also on a number of parameters and setting conditions - on the width and symmetry of the analyzer window setting, on the radiation energy, on the amount of scattered radiation, on the frequency of the pulses, on the collimator used. It can also change over time due to instabilities in the detector and electronic circuits. If a sufficient number of accumulated pulses in the image (information density) is not achieved, statistical fluctuations may also appear as inhomogeneity - the correction matrix must therefore be recorded with the highest possible number of pulses, min. 10,000 imp./pixel. These important aspects of measuring and correcting inhomogeneity are discussed and documented on experimental scintigraphic images in the section "Testing and calibration of camera image homogeneity", section "Dependence of inhomogeneity on physical conditions".
   Scintigraphic digital cameras have both of these operations 1. and 2. covered by the procedure calibration homogeneity and tuning of photomultiplier tubes.
Regular testing of the homogeneity of the gamma camera imaging 
Virtually all disturbances and anomalies in the imaging properties of the gamma camera are most sensitive manifest in homogeneity of the field of view. To ensure high-quality scintigraphic imaging, it is therefore necessary to perform regular homogeneity testing - and of course also after each electronic intervention in the circuits of photomultipliers, amplifiers and ADCs. In case of degraded homogeneity, it is necessary to adjust or recalibrate the photomultipliers
(tuning), updating the correction matrix, in case of gross abnormalities electronic intervention or repair. It is useful to archive the results of homogeneity testing for a long time, or to plot them graphically, for the analysis of the time trend and to reveal the causes of possible deterioration of homogeneity.
  How homogeneity measurements are performed using point and planar sources (and recommended testing intervals) is described in the work "Phantoms and phantom measurements in nuclear medicine", section "Testing and calibration of camera image homogeneity".

Linearity of gamma camera imaging
Another parameter of scintigraphic image quality indicates whether the spatial scales and proportions in the object are displayed faithfully and without distortion - linearly. It is therefore the ability of the camera to display the distribution of radioactivity without positional distortion, to display the line source as an exact line. Special phantoms are used to assess (and possibly quantify) the linearity of the scintigraphic image, in which a regular geometric structure of the radioactivity distribution is realized. It can be either a system of a larger number of regularly quadrangularly distributed point sources, or a system of linear (straight line) sources
(so-called bar-phantom, mostly transmission). The most perfect phantom for the analysis of linearity of scintillation camera imaging is the cartesian linear grid - it is described in the work "Phantoms and phantom measurements in nuclear medicine", passage "Analysis of linearity of gamma camera imaging".
   The scintigraphic image of such a regular geometric structure should also show geometric regularity. Possibly nonlinearity of imaging wil be reflected in this image as distortion and irregularities in the geometric arrangement. We can monitor them either visually or evaluate them quantitatively :

Linearity of scintigraphic imaging (spatial)
The linearity of the scintigraphic image is characterized by the maximum deviation of the scintigraphic image of the linear distribution of radioactivity from the exact linear form :
L
[mm] = max (X - X
lin ) ,
where X are the actual coordinates in the image and X
lin are theoretical coordinate values corresponding to the exact linear course .

Sometimes linearity is also expressed as a percentage, ie L = 100[%] .max (X - X lin ) / Xlin. Linearity is analyzed in two "X" and "Y" directions perpendicular to each other. As with resolution and homogeneity, linearity is also given for the whole and central field of view, or in addition to the total (absolute) linearity, the differential linearity is also given. The spatial linearity of the display should be better than about 4 mm for high-quality and correctly adjusted cameras.
Linearity
<---> Homogeneity
 
The linearity of the image and the homogeneity of the sensitivity of the field of view of the scintillation camera are closely related. Irregularities in the efficiency of registration of scintillations from different places of the scintillation crystal of the camera by a system of photomultipliers will be reflected in the image as geometric nonlinearity and at the same time as inhomogeneity in the density of registered pulses. It can be said that under normal circumstances the nonlinearity of the image is the main source of image inhomogeneity. Deviations in the regular arrangement and size of the holes and baffles in the collimator can also contribute to the inhomogeneity of the scintigraphic imaging, especially in the case of mechanical damage to the collimator.
   In practical testing of the properties (quality) of a scintillation camera, the linearity of the image is rarely determined, as it is difficult and, in addition, a small change in linearity (which would be difficult to demonstrate in targeted linearity measurements) is significantly more manifested in the inhomogeneity of the field of view. The only case of visually observable systematic nonlinearity is scintigraphic images with convergent and Pinhole collimators
(§4.2, section "Scintigraphic collimators").

Tomographic resolution, homogeneity and linearity
For all the above parameters of the classical scintillation camera, we had in mind the usual planar scintigraphic imaging. The quality of the tomographic image is basically described by the same physical parameters as in the planar image. However, the planar parameters of the SPECT *) scintillation camera detector cannot always be transferred directly to tomographic images of transverse sections, which arise from a complex reconstruction procedure from many planar images at different angles. Although the parameters of the camera detector are also decisive for the quality of tomographic images, some other physical and technical aspects also cooperate here.
*) Of course, there is no planar image for the PET camera, all parameters are measured in tomographic mode, on images of transverse sections - see §4.3, section "Spatial resolution of PET".
   The spatial resolution of the camera in the planar image is also decisively reflected in the tomographic image. The radial tomographic resolution in the transverse section image is approximately (1.1-1.3) times the total resolution of the camera, but it can be possibly aggravated by mechanical shifts of the center of rotation during the acquisition of SPECT examination. The resolution in the axial direction is given directly by the resolution of the camera with the collimator used (at a given distance of the displayed structure from the front of the collimator).
   Possible local defect in the homogeneity of the camera, it is projected as an annular artifact in the rotation of the camera in the cross-sectional image. The inhomogeneity in the cross-sectional image can then also be caused by the absorption of
g radiation (attenuation) depending on the depth of deposition of the respective radioactivity distribution in the tissue.
  Tomographic resolution and homogeneity are measured or evaluated (mostly visually) using special aids - phantoms, most often cylindrical in shape, containing tubes (line sources), various rollers and balls of various sizes, as well as free space for homogeneous distribution of the radio indicator. The most commonly used is Jasczak 's phantom. The phantom is filled with a radionuclide solution (usually
99mTc or 18F), its SPECT or PET scintigraphy is performed and the resolution and homogeneity are evaluated on the reconstructed images of transverse sections in the appropriate places in a manner analogous to the planar images. Tomographic phantoms are described in the work "Phantoms and phantom measurements in nuclear medicine", section "Tomographic phantoms".

Energy resolution and dead time of the camera detector
In addition to the above basic parameters - spatial resolution, homogeneity and linearity of images, which have a primary effect on scintigraphic image quality, for the scintillation camera are also assesed detection parameters, describing its properties in terms of scintillation detection and radiation gamma spectrometry. Although these parameters are secondary and auxiliary, they can indirectly affect the quality of the display. Their adverse changes may also indicate a malfunction of the scintillation camera.
The energy resolution
of a scintillation camera not only allows the separation of different lines of gamma radiation
(eg when simultaneously imaging two isotopes), but mainly determines the ability of the camera detector to separate Compton - scattered radiation g from direct non - scattered radiation (discussed above in the section "Spatial resolution of the gamma camera", passage "Compton - scattered radiation" , Fig.4.5.4). With the correct adjustment of photomultipliers, the total energy resolution of classic gamma cameras (Anger type) is about 9-12% (for new semiconductor CZT cameras it is about 5%).
Dead time
(sometimes called time resolution) is the time for which the detector processes the signal from the arrival of one detected quantum of radiation and is not able to register any other quantum. In scintigraphy, it can manifest itself to sources with high activity, ie with a high flux of photons of radiation
g - at a frequency of many tens of thousands of registered photons per second. The dead time of the camera leads to a violation of the linearity of the dependence between the activity in the source and the registered frequency of pulses in the image, which may distort the results of the analysis of the dynamics of the investigated processes. For quantitative dynamic studies, especially radiocardiographic studies, the need for a correction for dead time may arise. The important thing here is overall dead time of the whole system camera + computer. For older types of cameras, the total dead time was about 5 ms, for newer cameras it is already reduced to about 1 ms. Overloading the gamma camera with high radiation intensity also leads to deterioration of imaging properties - resolution, homogeneity, image contrast. All registered photons of all energies contribute to the dead time, ie not only the corresponding photo peak in the analyzer window, but also Compton scattered radiation.
Note: For details on the paralyzable (cumulative) and non-paralizable nature of dead time, as well as methods for its measurement and correction for dead time, we can refer to Chapter 2, passage "Dead time of detectors".
   Energy resolution and dead time are measured by the spectrometric methods described in Chapter 2 "
Detection and spectrometry of ionizing radiation", in particular in §2.4 "Scintillation detectors".

The issue of measuring the imaging properties of gamma cameras and the practical implementation of testing is discussed in a separate work
"
Phantoms and phantom measurements in nuclear medicine ".


4.6. Relationship between scintigraphy and other imaging methods
Scintigraphy is just one of several other imaging diagnostic methods used in medicine. Each of these methods has its uses, its advantages and disadvantages. In principle, diagnostic imaging methods can be divided into two groups :
¨ Anatomical-morphological , 
which show mainly the size and structure of tissues and organs. However, anatomical imaging lacks a functional aspect - it does not allow to recognize the biological nature of the displayed pathological structure.
¨ Functional-metabolic ,
which map blood circulation, metabolism, drainage, accumulation and other organ functions. However, functional imaging usually does not allow the exact localization of a pathological event or lesion in the organism - missing here the "background" of other structures, that are not displayed (because they do not have the appropriate "function"). In addition, functional imaging generally has a lower position resolution than anatomical imaging.
   To make a correct diagnosis, it is necessary to assess both anatonic-morphological and functional and metabolic symptoms of the disease. Only the combination of both mentioned images will make it possible to recognize the biological character of the depicted deposit and its exact location.
  To clarify the position and role of scintigraphy in the spectrum of other diagnostic methods, we will briefly compare the principles and diagnostic capabilities of the most important imaging methods.

X-ray imaging
The oldest and most frequently used imaging method so far is X-ray imaging (see §3.2 "
X-ray diagnostics"), whether it is planar or tomographic CT imaging. The penetrating X-rays generated in X-ray tube pass through the examined object (organism tissue), while part of the radiation is absorbed depending on the tissue density, while the remaining part passes through the tissue and is displayed either photographically or on a luminescent screen, or more recently by electronic detectors. This creates an X-ray image of the examined tissue, which is a shadow density image showing differences in tissue density. In certain cases, the contrast of the image can be artificially increased by applying suitable contrast agents. In addition, tomographic X-ray CT imaging provides images of transverse sections with high resolution (approximately 1 mm), from which a three-dimensional image of the examined area can be composed.

Ultrasonic sonography
Ultrasound is a mechanical (acoustic) wave of a substance (air, liquids, solids) with a frequency higher than the sound audible to the human ear, ie higher than 20 kHz. In matter, a wave propagates by oscillating its particles around an equilibrium position. In gases and liquids it propagates as longitudinal waves, in solids it can also have the character of transverse waves. In medical diagnostics, ultrasound with a frequency of 1-15 MHz is usually used. At higher frequencies, better spatial resolution can be achieved (due to the shorter wavelength), but more ultrasound is absorbed in the tissue.
   Ultrasound sonography or ultrasonography is based on the propagation of sound waves of high frequency (several MHz), ie ultrasound, in the elastic environment of tissues and its reflections on inhomogeneities. The speed v the propagation of a wave in an elastic medium is given by the relation v =
Ö(M/r), where M is the elasticity (Young's modulus of elasticity) of the medium and r its density (specific gravity). When an ultrasonic wave strikes an area with different density or elasticity - the acoustic interface, there are changes in the speed of propagation, refraction and reflection of the wave (related to the well-known Huygens principle). The reflected ultrasonic waves carry information about the presence of structures of different density and elasticity. Ultrasound sonography creates an image of these structures in the examined tissue by echolocation *) of reflected ultrasound waves. The reflected signals - acoustic echoes - correspond in their time sequence to the spatial distribution of reflecting structures in the investigated environment.
*) Echolocation is a way of obtaining information at a distance, where a sound is transmitted to the monitored environment, which is partially reflected from a possible object back to the place of transmission and is captured and evaluated there. From the time delay which elapses from the moment of sound transmission to the moment of receivivg the reflected wave (echo), the distance of the reflecting object can be determined. In nature, this principle is used by dolphins and bats for orientation and searching for food. In marine technology, so-called sonar is used, among other things to measure the depth of the sea. A radar works on a similar principle of radiolocation, which uses electromagnetic waves instead of sound - radio waves.
  Transmitting piezoelectric crystal of the probe, pressed into mechanical contact with the body surface
(good passage of waves into the skin is ensured by a layer of special gel), is periodically deformed by the action of an alternating electric voltage applied to its opposite electrodes, and this mechanical disturbance (vibration) emits an acoustic wave into the tissue. In water and tissue, sound waves propagate at an average speed of 1550 m/s. As ultrasound passes through the substance, it is absorbed, scattered, bent, and partially reflected back. Reflection occurs at the interface of tissues with different densities and elasticities, in which ultrasound propagates at different speeds - ie tissues with different acoustic impedances *). The reflected ultrasonic waves return and cause vibrations of the piezoelectric crystal (transducer) in the receiving part of the ultrasonic probe, which generates an alternating electrical signal of appropriate frequency, amplitude and time delay at the crystal electrodes, which is further electronically processed.
*) The so-called specific acoustic impedance is important for the propagation of ultrasound in tissue, which is the product of tissue density and ultrasound speed: Z = r .v = Ö(M. r). It gives the specific "wave resistance" in the propagation of ultrasound in analogy to Ohm's law of electricity. The acoustic impedance Z defined in this way is somewhat analogous to the reactance in electronics (capacitive or inductive). Viscosity the environment leading to the absorption and attenuation of ultrasound is analogous to the active ohmic resistance. The greater the difference in acoustic impedances, the greater the intensity of the reflections - echogenicity.
   The probe attached to the body surface transmits short (millisecond) ultrasound signals at fast regular intervals, the electronic receiving probe records the reflected signal ("echo") and the electronic apparatus evaluates the time and position differences of the transmitted and reflected signal and creates an image of structures on the screen according to their density and elasticity
(so-called echogenicity or acoustic impedance). Ultrasound images of echogenicity are displayed on most probes primarily in the form of a circular segment in polar coordinates (r, j) centered at the point of attachment of the receiving probe. The radial coordinate r - the depth in the tissue - is derived from the time delay Dt between the transmitted ultrasound signal and the reception of its reflection: r [mm] = 1.55 . Dt [ms] (at the average speed of ultrasound in the tissue approx. 1550 m/s). Angle j it is determined for simple devices with one receiving crystal by rocking turning of the probe (manual or motorized), for probes with more directional receiving crystals it is determined electronically. The brightness of the individual points of the sonographic image is modulated by the intensity of the received reflected ultrasound signal (this intensity should be corrected for the depth absorption of the ultrasound, see below), ie the echogenicity *) of the corresponding sites in the tissue. The sonographic image thus captures the spatial distribution of structures with different densities and elasticities in the examined tissue.
*) Formations that reflect ultrasound more or less than the surrounding tissue are called hypoechogenic or hyperechogenic. Higher or lower echogenicity is not in itself a "diagnosis", but it is an important feature by which the character of the examined area of tissue can be determined.
   Some technically advanced systems have a computer transformation of the image into the usual Cartesian coordinates, providing a more illustrative presentation. The rectangular view is provided by probes with a linear arrangement of a plurality of receiving elements.
   The signal coming from a greater depth is significantly attenuated by absorption in the tissue *) (double attenuation - transmitted and reflected signal), so for objective display, a correction of intensity for attenuation is performed, either time (longer signal reception time from greater depth) or computer.
*) For absorption of ultrasound in the environment occurs by the fact that due to the internal friction of the oscillating particles, part of the mechanical energy of the waves changes into heat. The rate of absorption of ultrasonic waves is determined by the exponential law I(d) = I0 .e - m .d , wherein I0 is the original (initial) intensity, I(d) the intensity at depth d , m coefficient of absorption. The absorption coefficient m depends on the type of substance (its viscosity) and on the frequency. In most substances, the attenuation by absorption is directly proportional to the square of the frequency.
   The great advantage of ultrasonography is the simplicity of its implementation, non-invasiveness and unpretentiousness for patients. The method is completely safe and harmless (ultrasound intensity reaches a maximum of 1 mW/m2), it does not burden the body with ionizing radiation. Therefore, in diseases related to morphological and anatomical changes, ultrasound examination is usually included at the beginning of the diagnostic chain. Ultrasonography is also widely used to evaluate the course of gravidity.
Doppler ultrasonography
Modern sonographic instruments also allow the analysis of the frequency of the received ultrasound signal: the frequency of the signal reflected from a moving object is slightly increased or decreased due to the Doppler effect *), depending on whether the object is moving toward or away from the receiver. The Doppler frequency shift of the reflected ultrasound can be then used to modulate a common echogenic anatomical image (color modulation is used) and thus obtain a velocity map of the movement of structures and fluid flow in the examined object. With the help of this so-called Doppler ultrasonography, it is possible in cardiological diagnostics (Doppler echocardiography) to detect, for example, movements of heart walls and valves, or jets of blood from under the heart valve during regurgitation. It is also possible to monitor the speed of blood flow in the venous system.
*) Doppler effect :
If the wave source moves of a certain constant frequency f
o towards the observer (receiver), this observer registers a higher frequency f than the source actually emits. Conversely, when the source moves away from the observer, the registered frequency is lower than the actual one. The difference between the actual fo and the observed f frequency (Doppler frequency shift) increases in proportion to the velocity V of the source relative to the observer: f = [1 + (V/v)] . fo , where v is the speed of propagation of a given wave. This rule also applies when the source of the received wave is the reflection of the wave from a certain moving object (including a flowing liquid). By measuring the frequency difference of the primary transmitted wave and the reflected wave ("echo") we can thus determine the speed of movement of the reflecting object.

Nuclear magnetic resonance - analytical and imaging method
Nuclear magnetic resonance (NMR) is a very complex physical-electronic method, based on the behavior of magnetic moments of atomic nuclei under the action of an alternating radio frequency signal in a strong permanent magnetic field. This originally analytical method was later improved and developed as an important imaging method .
Note: We have included nuclear magnetic resonance among nuclear and radiation methods, even though it does not contain any ionizing radiation. However, it is a method based on the findings of nuclear physics - the properties of atomic nuclei. A physical phenomenon called nuclear magnetic resonance - NMR, was investigated in the 1940s (F. Bloch, E.M.Purcell) and was initially used in chemistry as sample NMR spectrometry . In the 1970s and 1980s, NMR imaging methods also began to develop (pioneers were P.Lauterbuer, P.Manfield, A.Maudsley, R.Damadian, 1977) - see below.
    We will try to briefly outline the principles and methodology of NMR. However, due to the considerable principal and technical complexity of NMR (only scintigraphy can partially compete with it), we must observe the maximum brevity ...
Physical principle of NMR
Phenomenon of nuclear magnetic resonance it can generally occur during the interactions of atomic nuclei with an external electromagnetic field. Each nucleon (proton and neutron) has its own "mechanical" angular momentum - spin (nucleons belong to fermions with spin 1/2, see §1.5 "
Elementary particles"). According to the laws of electrodynamics, this rotational angular momentum of nucleons creates - induces - its own elementary magnetic moment mp = 1.41.10-27 J / T, equal to 2.79 times the so-called Bohr nuclear magneton *) - it is discussed in more detail in §1.1, passage "Quantum momentum, spin, magnetic moment", paragraph " Magnetic moment ". Due to the spins of their nucleons, atomic nuclei therefore generate a very weak magnetic field - they have a certain magnetic moment m . However, only atomic nuclei with an odd nucleon number have spin and magnetic moment, because the spins and magnetic moments of paired protons and neutrons cancel each other out - they are zero. The magnetic moment of the nucleus is formed by an unpaired nucleon - a proton or neutron. Magnetic resonance imaging can therefore be observed only in nuclei with odd nucleon numbers - especially hydrogen 1H, then in 13C, 15N, 19F, 23Na, 31P, etc.
*) Nuclear magneton
mN is a physical constant expressing the proton's own dipole magnetic moment induced by its spin: m N = e.h /2mp , where e is the elementary electric charge (proton, electron), h is the reduced Planck's constant, mp is the rest mass of the proton. In the system of SI units, its value is approximately mN = 5.05.10-27 J /T. It is analogous to the Bohr electron magneton me = e.h / 2m e, which, however, is 1836 times larger, in the ratio of the mass of the proton and the electron. It is interesting that even a neutron, although electrically uncharged, has a non-zero magnetic moment mn = -0.966.10-27 J /T somewhat smaller and of the opposite sign than a proton. It turns out that the magnetic moment of nucleons has its origin in their quark structure (§1.5., part "Quark structure of hadrons" and §1.1, passage "Magnetic moment").
Magnetic moments of nuclei in a magnetic field 
Under normal circumstances, due to the thermal motion of atoms, the directions of spins and magnetic moments of individual nuclei are chaotically "scattered", their orientation is random and disordered (Fig.3.4.4a), elementary magnetic fields cancel each other out on average, on a macroscopic scale the substance shows no magnetic properties
(we do not mean ferromagnetic substances, where it is the effect of electrons in atomic shells) . However, if we place the analyzed substance in a strong magnetic field (of intensity or induction B of the order of several Tesla), the magnetic moments of the nuclei are oriented in the direction of the vector B of this external magnetic field (at least partially).- the magnetic moment of the nuclei is parallel to the magnetic field lines (Fig.3.4.4b). The stronger the magnetic field, the more perfect this arrangement *). Outwardly, this results in non-zero magnetization vector M in the direction of the external magnetic field induction B. The magnitude of the magnetization vector is proportional to the strength of the external magnetic field B and the percentage of concordantly oriented mag. moments of nuclei in matter. A sufficiently strong magnetic field B is now mostly realized by means of a superconducting electromagnet, the winding of which must be permanently cooled by liquid helium (physical principles of superconducting magnets are briefly discussed in §1.5, section "Electromagnets in accelerators", passage "Superconducting electromagnets").
*) However, the extent of this arrangement is actually very small ! In commonly used magnetic fields 1-3T, for every 1 million hydrogen nuclei, only about 7-20 nuclei are on average in a state of uniform orientation. The vast majority of nuclei are as a result of thermal motion, it is oriented in different directions, including the opposite one (this is expressed by Boltzmann's law of distribution.) In this sense, it is necessary to take Fig.3.4.4b only as a symbolic scheme, which shows only those few nuclei that acquire concordant orientations.
  Since conventional material, e.g. water or tissue, contains about 1022 hydrogen nuclei per 1 gram, even a small excess of oriented nuclei provides a measurable magnitude of the magnetization vector and the radio frequency response signal.
Larmor frequency, radiofrequency excitation and relaxation 
In the magnetic field B, the nuclei (with a non-zero magnetic moment
m) behave as magnetic dipoles, which are acted upon by a pair of forces m.B . This will cause the core to rotate the axis of its magnetic moment around the direction B - it will perform a precessional movement (similar to the precessional movement of a gyroscope or children's "spinning top" around the vertical direction in the gravity field) by the so-called Larmor frequency
      
wL = g . B  , or    fL = g .B /2p ,
where g is the gyromagnetic ratio of the nucleus, which is the ratio of the magnetic moment of the nucleus and its "mechanical" moment of inertia [ rad · s -1 · T -1] . The precession movement occurs when the external magnetic field changes or the angle of the magnetic moment in this field changes and lasts as long as the mag. the moment does not stabilize in the rest position.
    If we send a short alternating electromagnetic signal into such a magnetically polarized substance by means of another coil (high-frequency - HF, or radio-frequency - RF)
(whose frequency resonates with the above-mentioned Larmor precession fL of a given type of nucleus in a magnetic field), the direction of the magnetic moment of the nucleus temporarily deviates from the direction determined by the vector B of the external magnetic field (Fig.3.4.4c) *). The deflection of the magnetization vector is caused by the magnetic component of the excitation RF pulse. The angle of this deflection is proportional to the amplitude (energy) of the RF pulse and its duration. The most commonly used RF pulses are 90° or 180°.
*) Fulfillment of the resonance condition: The nuclei are able to efficiently receive energy from an alternating electromagnetic field only if the Larmor frequency of the nucleus precession and the frequency of the electromagnetic pulse are the same. The preceding nuclei thus resonate with an electromagnetic pulse at a given Larmor frequency - hence the name "magnetic resonance".
    After the unwinding of the excitation, signal occurs relaxation
(at a constant rotation Larmor frequency) at which they emit electromagnetic waves with decreasing intensity until the magnetic moment of the spiral return back again in the direction B. These electromag. waves will induce alternating voltage in the receiving coils - "echo" radiofrequency signal **). This relaxation signal (sometimes referred acronym FID - Free Induction Decay) , has a sinusoidal course with exponentially decreasing amplitude (see below Relaxation times). It is a useful signal that carries information about the inner structure of the analyte. Frequency of this signal is equal to the above-mentioned Larmor precession and for a given force B of the external magnetic field is determined by the gyromagnetic ratio g of the nucleus, ie the type of nucleus, the amplitude of the relaxation signal is proportional to the concentration of nuclei of the given species- thus nuclear magnetic resonance can be used to analyze of the composition of substances : what elements and in a what concentration are contained in the sample. E.g. for hydrogen nuclei (protons) the gyromagnetic constant has the value g = 2.675.10-8 s-1 T-1 and in the magnetic field of induction 1Tesla Larmor's NM the resonant frequency is 42.574MHz, at 1.5T it is 63.58MHz - the area of radio waves (short waves) . For heavier nuclei is proportionally lower .
**) Phasing of a large number of nuclei : The NMR receiving coils are, of course, not able to detect the relaxation radiation of one or a few nuclei. To obtain a measurable signal, deexcitation of a large number of nuclei (> about 1012 ) is required, namely synchronously and at the same phase ! If phasing disruption occurs, the MNR signal drops sharply or disappears (cf. below "Relaxation times - T2").
    General note:
Quantum behavior:
For the sake of clarity, we have not yet explicitly included the quantum behavior of the magnetic moment, we considered its continuous behavior. The orientation of the magnetic moment vector of nuclei in a magnetic field actually acquires discrete quantum states - parallel (0°), perpendicular (90°) and antiparallel (opposite, 180°) with the direction of the vector B
  magnetic induction of an external magnetic field. The basic, energetically lowest state is parallel, while the perpendicular or antiparallel configuration has a higher energy- excited state. From the fundamental to the excited state of the magnetic moment, the nuclei pass by absorbing a quantum of electromagnetic energy, which must be exactly equal to the difference in energy between the two states. The respective frequency corresponds to the resonant Larmor frequency. During deexcitation, an electromagnetic signal of the same frequency is then emitted . The precession rotation of the magnetic moment of nuclei in a magnetic field is again just our model idea of how to clearly explain the behavior of nuclei in a magnetic field ...


Fig.
3.4.4. Nuclear magnetic resonance - simplified schematic representation.
a)
The magnetic moments of the nuclei in the analyte normally have chaotically scattered directions.
b)
By the action of a strong magnetic field B, the mag. moments of nuclei partially orient in the direction of the vector B .
c)
By sending a RF electromagnetic field, these oriented nuclei deviate from the B direction, eg by 90°. After switching off this RF field, a relaxation occurs, during which the deflected nuclei when its return at precession rotation will emit an electromagneic NMR signal with exponentially decaying amplitude.
d)
Simplified schematic diagram of NMR imaging equipment. For simplicity, only one radio frequency (RF) coil is drawn, which electronically switches alternately to transmit and receive modes; usually there are separate transmitting and receiving RF coils.
(ADC = analog-to-digital converter, DAC = digital-to-analog converter) .

Radio frequency coils
RF coils are a kind of "antennas" of the NMR system, that transmit excitation RF signals towards the analyte, or receive response RF signals from the relaxing nuclei in the analyte. In principle, the same coil can be used as the transmitting and receiving coil, which is electronically switched to the transmitting and then to the receiving mode
(as symbolically drawn in the diagram in Fig.3.4.4d). However, better detection of the response NMR signal can be achieved by using a separate receiving RF coil. Due to the relatively high Larmor frequency (tens of MHz), RF coils have a very simple design: they are formed by a loop of wire of circular or rectangular shape, which is placed close to the analyzed material (sample or area of interest in the organism). Sometimes they are suitably shaped (bent into a saddle or cylindrical shape) to achieve better homogeneity of the RF signal in the analyzed area.
  A short but very strong radio frequency alternating current, of high amplitude, is introduced into the transmitting coil in various time sequences, instantaneous power up to tens of kW. In the receiving coil, a response signal is then induced from the relaxing nuclei, on the contrary, with a very low amplitude (of the order of millivolts), which for further electronic processing must be significantly amplified in a narrowband high-frequency amplifier. For NMRI imaging (see below), special RF receiving coils of various sizes and shapes are used to tightly encircle the analyzed area - for imaging the brain, joints, spine, etc.
NMR spectroscopy and analysis
NMR spectroscopy
is performed in such a way, that the frequency of the excitation RF signal gradually increases, this signal intermittently supplies the coils in the transmitting mode, there is always a switch to the receiving mode and the intensity of the RF signal is measured - echo - transmitted by a sample placed in the magnetic field B
o during the back relaxation of the magnetic moments of the nuclei. The frequency at which the resonant maximum occurs, the Larmor frequency, determines the type of nucleus (the highest is for hydrogen - 42.6 MHz for B = 1Tesla), the intensity of the resonant maximum determines the concentration of the relevant atoms in the sample. All nuclei of one isotope, inserted into the same magnetic field, should resonate at the same frequency by themselves. However, if the atoms of these nuclei are part of chemical compounds, the distribution of electrons in their environment differs and these electrons cause electromagnetic shielding of the nuclei. The effective magnetic field acting on the nucleus is then no longer Bo, but B = Bo . (1- s), where the shielding factor s , describing the shielding intensity, slightly depends on the chemical composition of the analyte. This change in the effective magnetic field causes a so-called chemical frequency shift in the NMR signal spectrum .
    Another effect affecting the fine structure of the NMR spectrum is the mutual interaction of the nuclei of neighboring atoms mediated by valence electrons. As a result of these interactions, the splitting of the resonant maxima of the studied nuclei is observed into 2-4 lines at a distance of about 20 Hz - there is a multiplicity of signal .
    Detailed analysis of frequencies, intensities and multiplicities in the NMR spectrum can therefore provide information on the chemical composition and structure of organic and inorganic substances. Modern NMR spectrometers are computer controlled, and the induced NMR signal is analyzed using a Fourier transform .
Relaxation times
After switching off the high-frequency excitation field, the deflected nuclei relax in the magnetic field - they return in a spiral path to the original equilibrium state in the direction B
o (which we refer to here as the "z" axis), which is observed in the receiving coil as a free reverberation of the induced RF signal with an exponential decrease in amplitude. The rate of this relaxation (or fading time) is influenced by the interaction of nuclear spins with surrounding atoms and the mutual interaction between nuclear spins. The NMR signal also encodes information about the surrounding atoms and molecules - information about the chemical composition and structure of the substance. The decay time of the resonant signal is characterized by two relaxation times T1 and T2 .
    Relaxation time T
1 , sometimes called spin-lattice (the name comes from the original use of NMR for the analysis of solids with a crystal lattice) , represents the basic time constant of relaxation of magnetic moments of nuclei from the deflected position to the equilibrium position in the direction of the permanent magnetic field. It captures the speed at which the deflected core releases energy to electromagnetic waves and the environment during relaxation, while the longitudinal magnetization in the z-axis direction returns to the original value of Mo according to the exponential law: MZ = Mo.(1 - e -t / T1) . It is defined as the time, during which the longitudinal magnetization at relaxation reaches (1-e)- times the original value Mo, whereby the signal drops to 63% (if the excitation of the magnetic moment of the core by 90° was performed).
    The relaxation time T
2 , sometimes called spin-spin, expresses the time constant with which, due to the mutual interaction of spins and magnetic moments of adjacent nuclei, leading to the dephasing of the precessional motion of magnetic moments, the magnetization decreases in the transverse direction x-y: M XY = Mo XY e - t / T 2 . T2 is defined as the time during which the transverse magnetization MXY decreases e-times.
    Note: The receiving coil in the MRI actually detects a shorter relaxation time marked T2* after the excitation pulse has ended. In addition to the relaxation time T2, it is caused by a steeper decrease in the transverse component of the material magnetization due to small changes in the inhomogeneity of the magnetic field, leading to desynchronization. In MRI imaging, this phenomenon is usually negative, it can be corrected or eliminated in the so-called "spin-echo sequence" - see below.
    The relaxation times T1 and T2 are the result of the interaction of resonant nuclei with their surroundings and characterize the chemical properties and structure of the investigated material. In medical use, they are often significantly different for healthy and tumor tissue.
  In the most commonly used external magnetic field of 1.5 T, the relaxation times T
1 and T2 of water and some human tissues (in the physiological state) have the following approximate values :

Tissue type:   water   oxygenated blood non-oxygenated blood   fat     muscles   proteins gray matter brain white matter brain   liver     kidneys  
T 1 [ms] 4300 1350 1350 250 880 250 920 780 490 650
T 2 [ms] 2200 200 50 70 50 <= 1 100 90 40 70

Relaxation times are characteristic of different substances and tissues - they depend on the concentration of nuclei, temperature, size of molecules, chemical bonds. It can be seen from the table that, for example, hydrogen nuclei tightly bound in fat or protein molecules relax much faster than protons weakly bound in water molecules.
NMR imaging - MRI
The NMR method originally served as an analytical method for the composition and structure of samples. Advances in electronics and computer technology in the 1970s and 1980s made it possible to use the NMR signal to create an image of proton density in an object under investigation. This created the NMR imaging method (NMRI - Nuclear Magnetic Resonance Imaging; the word "nuclear" is often omitted and the abbreviation MRI is used) - Fig.3.4.4d.
    In order to be able to detect NMR signals separately and locally from individual places of the examined object (organism or tissue) and use it to create an image , it is necessary to ensure spatial-geometric coding of coordinates in the examined object. This can be achieved by superimposing an additional gradient magnetic field in the direction of the X, Y, Z axis on the main constant homogeneous field B
o. These gradient magnetic fields in the direction of each X, Y, Z axis are generated by a respective pair of gradient coils. By changing the gradient magnetic field, we achieve that the magnetic resonance will always occur in a different place of the examined object. By this gradient magnetic coding of spatial coordinates we can then perform NMR imaging.
Gradient coils 
are "additional" electromagnets located in suitable places inside the main strong electromagnet. They are wound with copper wire or metal tape, dimensioned for relatively high currents of tens or hundreds of amperes. Gradient coils are supplied in pulse sequences with a relatively strong current (approx. 500A) from electronically controlled sources, which allow fast and accurate setting of the strength and direction of the excited magnetic field - an additional gradient field. They produce gradients in the range of about 20-100 mT /m. In order for MRI imaging not to take an enormously long time, the rate of gradient changes needs to be relatively high - it reaches about 100-200 Tm-1 .s-1; it requires a certain voltage (approx. 50-300V) to overcome the inductance of the gradient coils - the power supplies of the gradient coils are relatively robust (power). Strong current surges in the gradient coils when interacting with the magnetic field cause mechanical vibrations, which causes considerable noise during MRI. Longitudinal gradient coils (in the Z direction ) have turns wound in the same direction as the main coil, X (gradient in the left-right direction) and Y (gradient in the up-down direction) are formed by saddle-shaped coils with vertically wound turns.
    Note first the longitudinal gradient field Bz(z) in the Z direction. His superposition with the main mag. field Bo causes the actual "local" value of the magnetic field B = Bo + Bz(z) to depend on the z coordinate : B = B(z). If we send a high-frequency pulse of a certain frequency f to a sample placed in this slightly inhomogeneous gradient magnetic field, the magnetic resonance signal will be transmitted by atomic nuclei only from a thin layer of the sample with coordinate z , for which the resonance condition f = g .B(z) /2p is satisfied. By varying the frequency f of high-frequency excitation pulses, or the intensity of the longitudinal gradient field Bz, is changes the position of the layer, in which the magnetic resonance response signal is generated. In this way, information about the dependence of the spatial distribution of the density of the nuclei in the direction of the Z axis is captured - the electronic-geometric coding of this coordinate is achieved - the layer z .
  The representation of the spatial distribution of the density of nuclei in a given layer z in the transverse directions X and Y is then obtained by the action of another, transverse, gradient magnetic field in the direction of the X and Y axis, whereby the investigated layer decomposes into elementary volumes - "pixels", in which is determined intensity of the relaxation NMR signal, and also its decay times. By changing these gradient fields, data are obtained for individual sites in the z layer and their computer reconstruction yields a cross-section image of the proton density in the examined layer z (Fig.3.4.4d right). By electronic analysis of relaxation times of the NMR signal is also generated cross-sectional images in the relaxation times T
1 and T2 (referred to as T1 or T2 - weighted images). The set of cross-sectional images for different values of the z-coordinate then creates a 3-dimensional tomographic image of the investigated area in proton density and relaxation times in individual "voxels". Using computer graphics, it is then possible to create images of any sections of the examined area, which are brightly modulated in a wide range of shades of gray (from white to black), to distinguish the structure of tissues and organs.
    The basic subject of NMRI imaging is hydrogen nuclei - imaging of proton density and relaxation times. This is why NMRI is sometimes referred to as "hydrogen topographic imaging". The intensity of such an NMR image mainly reflects the amount of water at each locationin the examined tissue and the nature of the binding and distribution of hydrogen molecules in the cells and extracellular space, as well as the distribution of fat and proteins. Based on these structural differences, different tissues can be distinguished from each other in MRI images - such as water, muscle, fat, gray matter and white matter in the brain.
    In general, two basic information is captured locally in NMRI images :
1. Density distribution of nuclei producing nuclear magnetic resonance - most often the proton density PD of hydrogen in the tissue. PD images essentially capture the anatomical structure of tissues and organs, and are largely similar to CT X-rays, which map the electron density of tissues.
2. Distribution of relaxation times T
1 and T2 related to the chemical composition and structural state of the tissue in individual places. Such images are called T1 and T2 - weighted .
    About to what extent to which the proton density PD and the times T
1 and T2 will be represented in the resulting MRI image, - how and with what this image will be modulated - "weighted" - is determined by pulse sequences: time sequence of transmitted RF pulses and "echo" response signals (will be discussed in more detail below) .


Fig.3.4.5 MRI images of the brain (transaxial section, without pathology) in proton density, relaxation times T 1 and T 2 and in a special FLAIR sequence to suppress the water signal.
(MRI brain images were taken by Jaroslav Havelka, MD, head of the MRI RDG department at the University Hospital Ostrava )

Proton densities and especially relaxation times are different not only for different types of tissues (see table above), but also differ depending on the physiological or pathological condition of the same tissue. This makes NMRI imaging an important diagnostic method in medicine, including in the field of tumor diagnostics.
    
Note: As with X-ray diagnostics, NMRI also uses contrast agents to increase the contrast of images of certain structures (eg cavities or blood vessels), but not on a density but on a magnetic basis - ferromagnetic compounds, mostly based on gadolinium .
Pulse sequence in NMRI
In medical MR imaging, it is desirable to create images with sufficient high contrast between different tissue types so that the MRI radiologist can best answer the clinical diagnostic question. Optimal image contrast between different tissues with different densities and rexation times can be achieved by suitable excitation of magnetic moments of nuclei and subsequent measurement of their response MR signal: by setting parameters of pulse sequence - time sequence of transmitted electromagnetic excitation pulses RF
and subsequent measurements of the "echo" of the electromagnetic signal emitted by the relaxing nuclei. The first parameter here is the intensity (energy) of the transmitted radio frequency excitation pulse (RF), which determines the predominant angle of deflection (tilt) of the magnetization vector of the analyzed nuclei - 90° or 180°. The higher the excitation intensity radiated into the analyzed target tissue, the higher the percentage of reversal of the magnetic moment of the nuclei and the stronger the response signal and more time is required for relaxation. Another parameter is the time interval TR , in which we repeatedly apply individual radiofrequency excitation pulses. The shorter this interval, the less time there is for T1 relaxation. The third parameter is the time TE (echo time) between the excitation pulse and the detection of the response resonant signal. The longer this time, the less nuclei with a shorter relaxation time T2 will contribute to the measured resonant signal. The completely approximate values of the pulse sequence times for obtaining the basic types of MRI images at B = 1.5 T are :
    
PD: TR = 1000 ms, TE = 5-30 ms; T1 -weighted: TR = 10 ms, TE = 5-30 ms; T2 -weighted: TR = 1000-2000 ms, TE = 80-100 ms.
    In connection with these regularities, several significant sequences of transmission of excitation radiofrequency pulses and subsequent detection of response relaxation signals have been developed (sometimes called "MRI techniques" in MR jargon ) :
-> Saturation - recovery sequence in which 90° RF pulses are transmitted at regular intervals. Upon arrival of each RF pulse, the magnetization vector rotates 90° and relaxation begins with different times T1 in different tissues. When another RF pulse arrives, the z-component of the magnetization will be different in different tissues. With a suitable repetition period TR of excitation RF pulses, we can set the optimal contrast of the desired tissues at times T1 . This simplest MRI technique is now rarely used, it has been replaced by the inversion-recovery sequence below, providing higher contrast.
-> Spin - echo sequence consisting of a 90° RF pulse followed by a 180° RF pulse. After the magnetization vector has been flipped into the xy plane due to a 90 ° RF pulse, T2 (resp. T2 *) relaxes, during which phasing occurs. However, the subsequent 180° RF pulse has a "refocusing" effect - it flips the individual spins in the xy plane by 180 ° and the spins are phased again. The result is an echo signal in the receiving coil, the amplitude of which depends on the relaxation times T 1 and T 2 of the tissue (unfavorable T2 * does not apply here, because the effect of magnetic field inhomogeneity on phasing is eliminated by 180 ° pulse phasing) . The character and contrast of the display can be adjusted using the times TR and TE. With short TR and short TE we get T1-weighted image, long TR and short TE provide a proton density image, long TR and long TE provide a T2 -weighted image. Due to this variability of imaging options, spin-echo is the most commonly used MRI technique.
-> Inversion - recovery sequence, consisting of a sequence of 180° and the following 90° RF pulse. The initial 180° pulse inverts the magnetization vector to the opposite, after which T1 relaxation takes place . With a time interval TI - inversion time , a 90° RF pulse then follows, which flips the magnetization vector into the xy plane. A RF signal dependent on T1 is detected in the receiving coilrelaxation time of the displayed tissue. The contrast of the image can be adjusted appropriately using the TI time. A significantly more contrasting image can be achieved than with the saturation recovery technique.
By a special setting of the time T1 = T
1 .ln2, the suppression of the image of the tissue having this relaxation time T 1 is achieved . By setting the short inversion time TI (approx. 140ms with a 1.5T magnet) - the so-called short time inversion recovery STIR - the suppression of the fat signal is achieved in the image . Conversely, by extending the time TI (to about 2600ms) - fluid attenuation inversion recovery FLAIR - we can achieve suppression of the water signal. Other fine details and anomalies in the structure of the examined tissues can then be better assessed on such "cleaned" images.
-> Gradient - echo sequence begins with a 90° RF pulse (which tilts the magnetization vector to the xy plane), after which a magnetic field gradient is applied. The nuclei in adjacent atoms will thus show a precession with a slightly different Larmor frequency, which will cause spin phasing. The application of the second mag. gradient with the opposite sign, which rephases the spins and at this point the echo is measured. Used to obtain a T2 -weighted image.
-> .......... sequence ............ ? add more sequences? ........... ? complete the picture of the graphic sequence diagram? ...
    Computer analysis of MRI images obtained with appropriate sequences
(mentioned above) can create special image modulations - such as water or fat signal suppression images . Other special sequences are used for functional MRI (mentioned below) :
-> Susceptibility weighted imaging ( SWI ) shows tissues with slightly different magnetic susceptibility. It uses an extended gradient-echo sequence for display in T2*. Its main variant is Blood oxygenation level dependent (BOLD), see fMRI below .
-> Diffusion weighted imaging (DWI) shows the diffusion of water inside tissue elements, manifested by Brownian motion of molecules. Using a spin-echo sequence with the application of 2 gradients, a subtle effect is registered, in which Brownian-moving water molecules show a different phasing-phasing relationship when reversing the mag. gradient; this leads to a slightly weaker T2 signal.
MRI Magnetic Resonance Spectrometry MRI
Magnetic resonance imaging (MRS) can be supplemented by the magnetic resonance spectrometry (MRS) described above, which enriches this examination with additional physiological information. Chemical analysis is performed here by analyzing the chemical shift of the Larmor frequency imaging structures in-vivo, eg choline or lipid levels. Chemical shifts are very fine, so this method is demanding not only in terms of signal analysis, but also requires high intensity (recommended at least 3 T) and homogeneity of the magnetic field.
Functional magnetic resonance imaging - fMRI 
Magnetic resonance imaging may be a suitable method for non-invasive imaging of the function of various tissues and organs (along with "molecular" imaging in nuclear medicine - .....). So far, fMRI has found application mainly in functional brain imaging, mapping neuronal activity . Neurons
(which do not have internal energy stores) they need to get sugar and oxygen quickly for their increased activity. The hemodynamic response to this need causes an increase in blood perfusion at a given site, but mainly a greater release of oxygen from the blood than inactive neurons. This leads to a change in the relative levels of oxygenated oxyhemoglobin and non-oxygenated deoxyhemoglobin in the blood at sites of neuronal activity.
    In this respect, two basic methods of indirect mapping of neuronal activity are used :
- Local increase of perfusion at the site of increased neuronal activity - perfusion fMRI .
- Change in the ratio of oxygenated and non-oxygenated blood at the site of neuronal activity. The method is called BOLD fMRI (Blood Oxygen Level Dependent). Changes in the relative levels of oxy- and deoxy-hemoglobin can be detected based on their slightly different magnetic susceptibility. Basic hemoglobin without bound oxygen (deoxyhemoglobin) has slightly paramagnetic properties, but when oxygen is bound to it (oxyhemoglobin), it behaves slightly diamagnetically . If more deoxyhemoglobin accumulates at a certain site in the brain tissue, a slightly stronger MRI signal is obtained from it than from the sites where deoxyhemoglobin predominates.
  MRI functional imaging of the brain is performed after neurological activation , either motor
(eg movement of fingers) , visual, linguistic or cognitive.
The physical-electronic implementation of NMRI 
NMR imaging is the most complex imaging method. The operation of the device for NMR imaging is electronically very complicated and demanding, so it must be controlled by a powerful computer with sophisticated software - Fig.3.4.4d. In the multiplex mode, the process of transmitting a sequence of radio frequency pulses, modulation of gradient magnetic fields, sensing and analysis of relaxation signals of magnetic resonance, reconstruction and creation of the resulting images, as well as a number of other transformation and correction procedures are synchronously controlled. Since these are harmonic (sinusoidal) waveforms, scanning and reconstruction are performed using Fourier analysis - in the frequency so-called K-space. It is a set of matrices defined in the memory of the MRI evaluation computer, into the individual elements of which the frequencies, amplitudes and coordinates of MRI signals are recorded. From these "raw" data, the resulting MRI images are created using Fourier transform and other analytical methods.
    Note: Electron paramagnetic resonance (EPR) is based on a similar principle as NMR. The magnetic moments of the electron shells of atoms are used here .........

Thermography
Thermography is a method of imaging the temperature distribution on the surface of analyzed objects. The medical use of thermography is based on the fact that some pathological events in the body are accompanied by changes in temperature (eg the inflammatory process by raising the temperature), which are also reflected on the surface of the body at a location above the lesion. Thermographic imaging can be performed in two ways :
¨ Contact thermography
using liquid crystals. Liquid crystals are substances that behave mechanically as liquids, but optically as solid crystalline substances (they appear optical anisotropy). Thermography uses the properties of some liquid crystals that, depending on the temperature, they color differently (thermo-mechanical changes are reflected in the the interference of incident light). With a suitable composition, liquid crystals can be prepared, which make it possible to display different temperature ranges in color. The liquid crystals were formerly painted directly onto the skin with a black-backed. They are now coated on a special flexible foil, which attaches to the skin.
¨ Infrared electronic thermography
by scanning infrared radiation from the surface of examined bodies with a special video camera sensitive to infrared radiation. Each body of non-zero temperature emits infrared radiation (thermal radiation) - electromagnetic waves of a continuous spectrum with a wavelength greater than visible light. Its intensity is greater the higher the surface temperature; as the temperature rises, the average wavelength decreases. It is used in industry and construction - for example, in the infrared image of a heated building, the places of poorer insulation with higher heat leakage are clearly shown.
   Thermographic imaging of the body surface, obtained with a sensitive infrared camera, may show areas with abnormally elevated temperatures (differences may be less than tenths of °C), which may indicate inflammatory or tumorous process in the tissue lying beneath this precinct. Alternatively, low temperature areas may indicate a perfusion disorder, perhaps due to occlusion of a blood vessel (eg, venous thrombosis).

Electroimpedance imaging
Certain information about the properties of tissues can be determined by sensing the local electrical conductivity, resp. impedance of examined tissues. A weak electric current is introduced into the tissue by means of electrodes placed on the skin in the vicinity of the examined area, and also by means of electrodes the distribution of electric potentials on the surface is sensed. From these data it is possible to reconstruct the spatial distribution of local tissue impedance - electroimpedance image. A different electrical conductivity is observed in the tumor tissue from the surrounding tissue. This method is trying (so far rarely) in mammography.

Complementarity of methods
From this brief overview of the principle and statements of several different imaging methods, we see that each of the methods looks at the examined tissue or organ from a different "viewing angle" - it examines a different aspect of morphology or function. In other words, it can be said in general, that the individual imaging methods are complementary to each other - in certain aspects they complement and compose a diagnostic "mosaic", which is then interpreted by an experienced clinician into the final diagnosis, from which the appropriate method of therapy is derived.

Status and role of nuclear medicine
Scintigraphy does not provide images with as high a resolution as CT, it does not recognize density, temperature or mechanical consistency of tissues. The main "floor" of nuclear medicine is the non-invasive imaging and quantification of structures and processes in the body, which are characterized by a specific function and metabolism, which can be "traced" by a suitable radio-indicator and imaged by external detection of gamma radiation. The degree of local accumulation of radiopharmaceuticals depends on the intensity of local metabolic and functional processes in organs and tissues. Possible anomalies and malfunctions can be located and quantified using scintigraphic imaging. Disorders of function in many cases precede structural disorders - anatomical and morphological. Therefore, pathological events can be detected by nuclear medicine methods sometimes earlier than other imaging methods
(typical examples are bone metastases of breast, prostate or lung cancer). The perfusion of tissues or organs and the dynamics of blood flow through individual parts of the heart and blood vessels can also be analyzed very well by nuclear medicine methods.
  Scintigraphic diagnosis of tumorous diseases is very important through targeted uptake of specific radiopharmaceuticals (especially monoclonal antibodies), which can be effectively followed by biologically targeted radionuclide therapy - a theranostic approach.

Fusion of PET and SPECT images with CT and NMRI images
In this §4.6 "Relationship between scintigraphy and other imaging methods", it was continuously discussed how the individual diagnostic methods complement each other in creating a comprehensive picture of the healthy or pathological condition of individual organs and the whole organism. The main goal is to combine anatomy with physiology, or in other words function with morphology, in order to better clarify the location, biological character and origin of pathological foci and abnormalities - assign the foci shown on the scintigram to specific anatomical structures in the organism. In the field of imaging methods, in such multi-modalities examinations are scanned and compared the
CT , SPECT , PET , NMRI and sonography images.
   Scintigraphic images provide important information about the functional status of tissues and organs, but are usually unable to provide sufficient anatomical information about the exact location of pathological abnormalities (lesions) imaged scintigraphically. Radioactivity does not enter the surrounding anatomical structures
(eg skeletal), which do not capture the radioindicator and are not visible in the scintigraphic image. For a better and clearer comparison of the character, size and location of the displayed structures, it is optimal to perform a simultaneous display PET+CT images, or SPECT+CT (eventually +NMRI), into a single suitably color-modulated image - so-called image fusion. Individual images are overlaid in various color combinations. In these images, where it is possible to continuously modulate the percentage of individual images (from what percentage will one image be projected into another), we can observe the correlation of physiological and anatomical-structural information. This will make it possible to accurately locate scintigraphically displayed lesions in terms of space and anatomy.
   These mergers often encounter the problem, if images from different modalities were captured at different times, at different display scales, and with different geometric configurations of the patient relative to the imaging device. Sophisticated computer graphics programs are able to perform affine and conformal transformations of images to correct geometric effects - scaling and relative position of displayed structures
(translation, rotation, reorientation, enlarging or reducing, cross-correlation) and achieve a relatively good "matching" of images, but some deviations from the exact overlap of the corresponding structures may persist. For this process of geometric alignment of images, the somewhat misleading name of registration or normalization of images is sometimes used. Procedures of this kind are applied in a number of areas of computer image processing (eg in panoramic photography, cartographic imaging, astrophotography).
Hybrid tomographic systems - combination of PET+CT, SPECT+CT, (PET+MRI) 
In order to eliminate these problems, as well as to operatively and quickly achieve comprehensive diagnostics, an effort was made to combine some imaging methods into one device. Imaging device manufacturers have developed so-called hybrid systems, combining pairs of PET+CT or SPECT+CT devices (a combination of PET+MRI is also mentioned below). These combined systems have three basic advantages :

*) Geometric alignment of CT images with SPECT and PET images
When fusing functional scintigraphic SPECT or PET images with anatomical CT images, it is important that the structures shown in both modalities overlap geometrically - they appear in the same place of the image. To calibrate this precise harmonization of mutual overlapping position of the displayed SPECT <-> CT or PET <-> CT structures on hybrid instrumenrs, is used scintigraphic + CT imaging of point sources filled with mixture of radionuclides (
99mTc, or 18F) and the contrast agent ("Phantoms and phantom measurements", part "Tomographic phantoms for SPECT, PET, CT", passage "Geometric alignment of images CT images of SPECT and PET"). This process geometric alignment of the images are sometimes (not very accurately) called spatial normalization and registration of images.
   Hybrid combination PET/CT in recent years, it has become a standard feature of PET-nuclear medicine workplaces, and its benefits are undeniable, especially in the field of tumor diagnostics. Separate PET cameras, without CT, are no longer manufactured.
   The combination of SPECT+CT is also useful for refining localization diagnostics, although the proportion of tumor imaging in SPECT is lower. In the SPECT/CT combinations is sometimes used the simpler and cheaper "low-density" ("low dose", "non-diagnostic") localization CT, but this solution is not entirely optimal; for greater versatility are prefered hybrid combinations with full diagnostic multi-slice CT (which, by the way, can be operated even in a low-dose mode, by reducing the anode current). In many cases, the combination of SPECT/CT helps to refine the diagnosis by physiological-anatomical correlation.
Hybrid combination of PET + NMRI
 
Nuclear magnetic resonance imaging (NMRI) provides, compared to CT, better soft tissue resolution, which is particularly advantageous in oncological diagnostics. In recent years, therefore, efforts have been made to develop a hybrid combination of PET/NMRI imaging devices. However, the direct combination of "classical" PET and NMRI technologies into one system faces a major problem: the strong magnetic field of the NMRI superconducting electromagnet affects the movement of electrons between dynodes in photomultipliers of ring scintillation detectors in PET
(Lorentz force acts on electrons); in a strong magnetic field, the photomultipliers stop working. However, technologies have been developed to combine PET with NMRI :
¨ Semiconductor photodetectors. Instead of conventional photomultipliers, APD (Avalanche Photo Diode) semiconductor photodiodes or SiPM semiconductor photomultipliers are used in the PET ring detector (see §2.4, section "Photomultipliers"). These photodetectors are not sensitive to magnetic fields.
¨ Multipixel fully solid-state detectors (e.g. based CZT) of annihilation photons (instead scintiblokù BGO/LSO with conventional photomultipliers or SIPMA). In addition to better detection efficiency and the spatial resolution can thereby be achieved coincidence somewhat shorter time (for better TOF). The advantage of semiconductor detectors is also their independence on the magnetic field, which just allows the use in hybrid systems of PET/MRI.
Note: Initially, some provisional suboptimal solutions were proposed :
- NMRI magnetic field modulation, which would turn on only while magnetic resonance imaging was acquired, while it would be turned off when capturing PET images.
- Optical fibers - light scintillations from annular PET scintillators would lead to photomultipliers located outside the NMRI magnetic field using optical fibers.
These bizarre makeshift suggestions have not prove, did not work and were not used in practice. The only real possibility of a hybrid PET/MR combination is the use of magnetically independent semiconductor photodectors, or better directly semiconductor gamma detectors.
   A hybrid PET/MRI combination is sometimes abbreviated as mMRI - Molecular Magnetic Resonance Imaging: MRI provides imaging of morphological and functional details of tissue, PET shows tissue metabolism at the molecular-cellular level. The integrated PET/MRI system allows, in some cases, more accurate identification and determination of the extent and characteristics of malignancies, which can help plan effective treatment and eventually its effect. The system can also be used in neurology and cardiology.
   However, the combination of PET/MRI also has some disadvantages, due to which it cannot yet function as a routine alternative to PET/CT (so far it is more of a specialized device) :
- Relatively long magnetic resonance imaging time, which limits the number of patients due to the work shift and the short half-life of 18-F; it also causes blurring of images in the chest and abdomen with breathing movements. CT is incomparably faster, practically does not prolong PET examinations and provides sharp images of moving organs.
- Common contraindications MRI (pacemakers, metal implants, stents), while CT has no contraindications (other than PET).
- MRI images do not yet provide accurate density maps for the exact correction of PET images to attenuation g radiation in tissue, as provided by CT images.
- Significantly higher purchase price of equipment and high operating costs of MRI, compared to CT.
   The argument that MRI is a non-radiation method with zero radiation dose is not significant here. PET gammagraphy alone is loaded with a relatively higher dose, and due to the composition of patients, who will usually be treated with radiotherapy with many times higher doses, the dose from CT imaging is irrelevant.
   Therefore,  installing a hybrid PET/MRI combination as the first or only PET device in a complex oncology center is not entirely optimal. A more suitable variant is the currently proven PET/CT combination; for the extension of complex diagnostics supplemented by a high-quality nuclear magnetic resonance device (3 T magnet) in a separate room or workplace, using computer fusion of PET+CT images with MRI images. In larger workplaces, even a hybrid combination of PET/MRI can be successfully used for some indications.
   A hybrid combination of imaging diagnostic and radiation radiotherapy technologies is discussed in §3.6 "Radiotherapy", section "Modulation of radiation beams".


4.7. Visual evaluation and mathematical analysis of diagnostic images
Diagnostic images - X-ray
(planar or CT), scintigraphic (planar, SPECT, PET) and magnetic resonance MRI can carry a lot of information about physiological or pathological anatomical-morphological situation and structure of tissues and organs, their function and metabolism, the presence of abnormalities and pathological lesions. To obtain this important diagnostic information - image evaluation - there are basically two ways to proceed :
¨
Visual evaluation 
by an "experienced eye" of an erudite expert in the field of nuclear medicine, X-ray or MRI diagnostics. This is the basic method of evaluation
(and before the era of digital imaging it was the only way...). An experienced radiologist can recognize a number of abnormalities, disorders, lesions on well-scanned images (with the necessary processing - brightness modulation, filtration, corrections). Such a description can then be an important guide for the detection and proper treatment of possible pathological conditions, as well as for the assessment of the response and effectiveness of the therapy.
¨ Quantitative processing , 
which with the help of mathematical-computer analysis provides quantitative parameters about the densities of various tissues and districts, the uptake of radionuclides and the function of the examined organs, on the course and rates of functional-metabolic processes
("molecular" imaging). The quantitative results obtained in this way complement and refine the visual assessment (eg the degree of metabolic activity of the visually recognized lesion), but they may also have their own importance for the assessment of functional processes in the organism (functional state of the kidneys, heart, liver). New special methods of filtering and computer image processing can also "pull out" and emphasize some details, indistinguishable in native images - and thus help visual evaluation.
Multifactorial statistical analysis of images, radiomics
In addition to the basic structural and functional information mentioned above, diagnostic images may also contain some additional information that is not directly visible (and does not result explicitly from quantitative analysis), but can in principle be extracted by special sophisticated computer methods of structure recognition and feature analysis in paintings. An example could be an analysis of the relationship between tumor size and shape (surface and volume), the degree of internal homogeneity or heterogeneity of displayed lesions and other semantic features in the images. It is also possible to analyze the topological shape and compactness
(such as the Hausdorf analysis of the contour dimension) and the relationship of the displayed structure with the surrounding tissues.
   Factors extracted in such a way do not provide any individually valid diagnostic information by themselves. Only when we confront them with statistical sets of a large number of evaluations of images of the same kind obtained in different patients with a reliably described diagnosis
(including genetic character), treatment method, response to therapy and overall outcome, can we - with some probability - reveal certain similarities and cerrespondences. This can potentially help to refine the diagnosis - by pointing to the possibility shown by this similarity - and possibly predict the response to the appropriate type of therapy and the further development of the disease.
   The necessary comparison databases can be created so that for each diagnosis considered in a standardized way  performs multifactorial analysis of images in many patients with the appropriate
(reliably verified) diagnosis and the quantified values of the extracted feature - factors from these images are gradually stored in special files, including the evaluation of statistical variance. In case of event. clinical use, then a particular patient performs display from which a multifactorial analysis of extracted features necessary (standardized same manner as in Comparative databases) and their values are compared with statistical methods - confront - with the respective "exemplary" values from the database. From "probability intersections" values of several factors can be inferred to a certain sub-type of pathology and possibly predict its behavior.
  These methods, based on "machine learning" and artificial intelligence, are still in the stage of experimental development, creation of necessary software, experimental compilation of factor databases from "sample" images (templates). For a set of these methods, using sophisticated image analysis methods in conjunction with statistical processing, the name radiomics has been used in recent years. Its perspective is to become a "bridge" between radiometric imaging and personalized medicine.

Mathematical analysis and computer evaluation of nuclear medicine <<-------click
   As a relic of the earlier structure of the monograph "Nuclear Physics and Physics of Ionizing Radiation", in which the chapter "Physical and Technical Problems in Nuclear Medicine" was included, the computer evaluation of scintigraphic studies consists of a separate set "Mathematical analysis and computer evaluation in nuclear medicine", which is then followed by a separate book "Complex computer evaluation of functional scintigraphic examinations on a PC - OSTNUCLINE system", describing specific procedures and algorithms, mathematical analysis and evaluation of scintigraphic studies of individual organs.
In this treatise "Radionuclide scintigraphy" the acquisition and evaluation of clinical scintigraphic examinations described in §4.9 "Clinical scintigraphic diagnostics in nuclear medicine".
  Furthermore, the work "
Filters and filtration in nuclear medicine" is related to this topic.


4.8. Radionuclides and radiopharmaceuticals for scintigraphy
  In order to be able to diagnostic imaging something at all with the help of scintigraphy, it is necessary should be to introduce into the body a g -radioaktive substance - radiotracer or radiopharmaceutical, whose distribution in various tissues and organs are then imaged. The indicator or tracking principle is used: radioisotopes behave chemically exactly like stable isotopes of the same element, but are "visible" through their radiation, which allows their monitoring in the system using ionizing radiation detectors, in the case of scintigraphy also imaging their distribution.
   Radiopharmaceuticals are special diagnostic or therapeutic preparations containing radionuclides, which are a source of radiation. A radioactive atom is incorporate in their molecules - by radionuclide we label a suitable compound that determines the pharmacokinetics according to our diagnostic or therapeutic requirements. The radiopharmaceutical is composed of two main parts :
× Carrier - specific biochemical substance providing pharmacokinetic targeting or directing to the desired site, tissue or organ which is to be displayed (or treated). The carrier is its own indicator of function, which actively or passively participates in the examined or therapeutic process in the target structure. In the simplest case, the carrier is water (saline), in which the radionuclide is dissolved and carried by the bloodstream, or air during pulmonary ventilation examination. However, most of them are more complex biochemically active substances - from inorganic salts, through cyclic hydrocarbons, chelates, dispersed colloidal particles, peptides, protein carriers, immunoglobulins, monoclonal antibodies, radiolabeled cells (erythrocytes or lymphocytes) - which are selectively taken up in target tissues, or pass through the bloodstream. The radionuclide carrier is selected according to the required diagnostic or therapeutic performance.
× Radionuclide bound to this carrier, ensuring by its emission of ionizing radiation "visibility" - signaling or indication positions of indicator molecules (carriers) in the organism - in our case display of its distribution. In therapy, it then causes biological effects on tissue cells. The binding of the radionuclide should be such as not to alter the biochemical properties of the carrier. Radiopharmaceuticals may also contain some stabilizing or antioxidant excipients.
   Radiopharmaceuticals are open radioactive emitters and, after application to the body, enter into various metabolic processes depending on their (bio)chemical structure.The chemical composition of the radiopharmaceutical determines its incorporation into kinetics or to certain metabolic processes, built-in radionuclide, by its radiation, then enables either external detection of the distribution of this substance
(in scintigraphy), or monitoring of its amount in the samples (biological fluids, mostly blood or urine). In the case of therapy, radionuclide radiation performs biological effects on the cells of the tissue in which the radiopharmaceutical accumulates (eg, it destroys tumor cells - §3.6, section "Radioisotope therapy with open emitters").
   The selected radiopharmaceutical should ideally accumulate only in the desired target areas. In practice, however, radiopharmaceuticals are to a greater or lesser extent also absorbed in other tissues and organs, or create a continuous tissue background. This undesirable pharmacokinetics should be taken into account when evaluating scintigraphic images, as well as when assessing side effects - radiotoxicity - in biologically targeted radionuclide therapy
(§3.6, section "Radioisotope therapy").
  Radioindicators in nuclear medicine are applied in small trace amounts, approx. 10
-9 -10-12 grams (pico- or nanomolar concentrations in tissues) and therefore on their own cannot affect function of the examined organs, nor can they cause any side or toxic effects on the organism. Therefore, due to the small and practically immeasurable amount by weight, radiopharmaceuticals cannot be dosed according to their weight [mg] - the weight of the "active substance", as is usual for drugs. Radiopharmaceuticals are dosed by the applied activity in [MBq].
Note: In this respect, radioindicators used in nuclear medicine differ significantly from contrast agents used in X-ray diagnostics. The X-ray contrast agent is applied in a relatively larger amount (up to tens of grams) needed to produce a sufficient contrast of X-ray absorption. There is a relatively high concentration in the blood and tissues, which due to the chemical composition of contrast agents (mostly iodine compounds) can significantly affect the function of the examined organs. Some contrast agents can have side effects or toxic effects, they can cause allergic reactions. In contrast, radioindicators used for diagnostics in nuclear medicine are biochemically safe and usually have no contraindications.
  A certain exception to this biochemical safety are radiopharmaceuticals based on murine monoclonal antibodies. They may have allergic reactions in a small percentage of patients, caused by the presence of so-called HAMA antibodies (Human Anti-Mouse Antibodies ); an adverse immune response to the preparation - production of human antibodies against murine monoclonal antibodies - may then occur. It is therefore desirable to perform a laboratory biochemical test for HAMA antibodies before using these radiopharmaceuticals, and its positivity should be a contraindication to the use of these products.
   Compared to other pharmaceutical preparations, radiopharmaceuticals have two other specifics :
«- Time-varying content of a substance carrying a diagnostic or therapeutic effect - the amount of radioisotope used decreases exponentially over time due to radioactive transformation (he rate of this decrease varies for individual radionuclides, depending on their half-life). This is associated with a short expiration time (which cannot be extended in any way, regardless of eg storage temperature).
«- Remote action - emitted ionizing radiation, especially penetrating gamma, can have biological effects - in this case undesirable - even outside the tissue where the drug was distributed (or even on another patient in the vicinity of the patient with applied radioactivity).
   For the use of scintigraphy in nuclear medicine, several
g- radionuclides are available (mixed b-g, EC, pure g, for PET then b+ with subsequent emission of annihilation g radiation) in the chemical form of a number of radiopharmaceuticals, enabling the study of various functional processes in the organism. Methods of production and physical (nuclear) characteristics of individual radionuclides are detailed in §1.6 "Radionuclides", where their measured spectra are also displayed. Here we will mention in particular the properties of the most important radionuclides used in nuclear medicine. For each such radionuclide, we draw its decay scheme, describe the methods of its radioactive transformations, types and energy of emitted radiation. Finally, for each radionuclide, we present the gamma-ray spectrum measured by a scintillation spectrometer with a multichannel analyzer - such a spectrum can then be observed in practice also on a scintillation camera; a more detailed semiconductor spectrum is also measured.

Radionuclides and radiopharmaceuticals for single photon gammagraphy - planar and SPECT
Radioiodine
131 I (+ 125I + 123I ) 
The first radionuclide used in clinical nuclear medicine was radioiodine
131 I (T1/2 = 8 days, b- with max. energy 606keV, the main energy g is 364keV), which is of key importance for the diagnosis and therapy of thyroid disease (§4.9.1 "Thyrological radioisotope diagnostics"). It is administered orally in the form of 131I- sodium iodide. For several years, radioiodine-labeled 131I-o-hippuran was also used for radionuclide nephrography and possibly renal scintigraphy, later displaced by 99mTc- labeled radiopharmaceuticals (see below).
   The radionuclide 131I is converted (according to the decay scheme in the figure on the left) by b- radioactivity to excited states of the daughter nuclide xenon 131Xe, which is already stable (non-radioactive). The dominant "channel" of beta-conversion is to an excited level of 364.5 keV (89%), which in 81% deexcites to baseline 131Xe and in 6% deexcites to a level of 80keV (which then deexcites to baseline). In 2% there is a decay to the excited level of 722keV, in 7% to the level of 637keV. One of the excited levels of 131Xe is the metastable 131mXe level with an energy of 164keV, which deeexcitates to the ground state of the 131Xe nucleus with a half-life T1/2 = 12 days. Only 0.38% of 131I decays occur at this metastable level, and in addition, its deexcitation is subject to internal conversion, so that only about 0.021% is emitted as 164keV gamma radiation. After about 14 days, a radioactive equilibrium is reached, when the activity of 131I is equal to the activity of 131mXe.


Fig .... Decomposition scheme and gamma-spectrum of radioiodine
131I

The spectrum of gamma radiation 131I is dominated by the main photopeak capturing the energy of radiation g 364keV. Towards higher energies, two weaker peaks, 637 and 723 keV, are visible. In the region of lower energies we also see weaker peaks 284 and 80 keV, at the very beginning of the spectrum the characteristic X-rays of Ka,b xenon 30keV (low-energy lines La,b 4-5keV on a conventional scintillation detector are not visible). The faint 164keV photopeak from the metastable 131mXe is not very noticeable, because it lies in the Compton scattering region of the main energy 364keV (interferes with the backscatter peak) - it is analyzed in " 131 I ".
   For in vitro radioimmunoassay (RIA, RSA) is then used radioactive iodine 125 I (T1/2 = 60 days, EC, X 27+31keV, g 35keV) ....
For scintigraphy is also used radioiodine
123 I (T1/2 = 13.1 hours, EC, g 159keV, X 27+31 keV), which has more advantageous physical properties for this purpose than 131I - more suitable energy g and the absence of b, which leads to a lower radiation load. Radiopharmaceuticals marked 123I are seldom used for scintigraphy of kidneys (o-jodhipuran), more often thyroid gland (NaI), heart (MIBG), as well as for scintigraphy of receptor systems in the brain - 123I-ioflupane, 123 I-IBZM (§4.9.8, part "Scintigraphy of receptor systems in the brain"). Compared to 99mTc, 123I has disadvantages in higher price, difficult distribution (short T1/2) and slightly higher radiation load.


Fig .... Decomposition scheme and gamma-spectrum of radioiodine
123 I

Technetium 99m Tc
The most important radionuclide for nuclear medicine is metastable technetium
99mTc (T1/2 = 6 hours), which is a pure gamma emitter (Eg = 140keV) and is obtained mostly by beta-decay of molybdenum 99Mo (T1/2 = 66 hours) in the so-called generator (see §1.2. "Radioactivity", part "Gamma radiation"). 99mTc is an almost ideal radionuclide for scintigraphy, on which basically the entire development of nuclear medicine in the 1960s and 1990s was based; has the following advantages :
×
1. A pure gamma emitter with a short half-life of 6 hours allows, without the risk of significantly increased radiation exposure, to apply to patients a relatively high activity of 99mTc (in the order of hundreds of MBq) required to obtain quality images in SPECT or dynamic scintigraphy.
Note:
After deexcitation of
99mTc, the 99Tc is formed in the ground state. It is also radioactive: it b- -transforms into a stable core of 99Ru (see Fig.4.8.2), but the half-life is very long here - 2.11.105 years. Since the activity of a preparation containing a given number of No radioactive nuclei is A = No . l, is the ratio of the activities of the parent and daughter radioisotopes in the ratio of their decay constants l, or the inverse ratio of their half-lives T1/2. The relationship between the activity of 99mTc and the activity of the formed 99Tc is thus given by the coefficient » 4.10-9. With an applied activity of the order of 100MBq 99mTc, the activity of the resulting 99Tc will be only about 0.4 Bq, which is practically zero (unmeasurable, well below the level of the natural radioactive background, eg 40K). Thus, from the point of view of nuclear medicine, the resulting 99Tc can be considered non-radioactive.
However, the opposite situation is in the field of nuclear reactors (see §1.3 "Nuclear reactions", passage "
Atomic nuclear fission"), where 99Tc, produced in significant quantities as one of the fission products of uranium, is a difficult component of nuclear waste with long half-lives, potentially hazardous to the environment.
× 2. Radiation g with an energy of 140 keV can be collimated very well and effectively detected in a thin large-area scintillation crystal of a gamma camera, which provides images with relatively good resolution and sensitivity.
×
3. 99mTc is easily obtained from a Mo-Tc generator (the physico-chemical principle of radionuclide generators is described in §1.2, part "Gamma radiation", passage "Radionuclide generators" and in §1.4 "Radionuclides", part "Production of artificial radionuclides", passage "Radionuclide generators"). These generators are mostly of the elution type. Molybdenum 99Mo is absorbed on a support (mostly Al2O3) in an "insoluble" oxide form in a "chromatographic" column. After the radioactive transformation of the 99Mo core into a 99mTc daughter core, the resulting technetium atom is released from an insoluble bond; combines with 4 oxygen atoms to form the anion 99mTcO4 - pertechnetate. This daughter product is soluble in water, whereby it can be separated from the starting molybdenum by washing with water - elution (Fig.4.8.1 left). Since the elution is performed with physiological saline containing a NaCl salt, the pertechnetate anions are immediately ionically bound to sodium to form sodium pertechnetate Na 99mTcO4-. In this chemical form we obtain technetium from the elution generator.


Fig.4.8.1. Elution
99Mo - 99mTc generator.
Left: Principle functional diagram of the elution generator. In the middle: Technical design of a sterile generator with an evacuated elution vial.
Right: Decomposition scheme of molybdenum
99Mo to technetium 99mTc, deexcitation to 99Tc and slow transformations to stable ruthenium 99Ru.

New types of sterile elution generators use an evacuated elution vial, into which, after "puncture" under atmospheric under-pressure, the saline solution is automatically sucked through a tube leading from the storage vial through the sorption column of the generator with 99Mo (Fig.4.8.1 in the middle). Within about 30 seconds, the vial is filled with 99mTc eluate.
   
In §1.2, part "Exponential law of radioactive decay", passage "Mixtures of radionuclides, decay series, radioactive equilibrium", the general equation of subsequent decay of radionuclides A(lA)®B(lB)®C (stable) was derived. If we apply this equation (multiplied by the factor lTc to get the instantaneous activity in [Bq] from the instantaneous number of nuclei) to our case of the Mo-Tc generator 99Mo(lMo=0,0105h-1)®99mTc(lTc=0,1155h-1)®99Tc ("stable") and taking into account that 87% 99Mo decays to a metastable excited level of 99mTc, we get for the time dependence of the immediate activity of the required technetium 99mTc relation: A99m-Tc(t) = 0,957.AMo(t=0) . (e-0,0105.t - e-0,1155.t ), where AMo(t=0) is the activity of 99Mo at time t=0 of the previous elution, time t is in hours. The activity of the 99Mo with time T varies according to the laws and decay AMo(T) = AMo(0).e-lMo.T = AMo(0).e-0,0105.T. Substituting this basic decomposition of molybdenum we obtain the resulting relationship for the instantaneous activity of the eluted 99mTc at time T from the delivery of the molybdenum generator and at time t since the last elution :
             A99m-Tc(T,t)  =  0,957.AMo(0).e-0,0105.T . (e-0,0105.t - e-0,1155.t )   ,
where A
Mo(0) is the activity of 99Mo at time T=0 of the generator supply, times t and T are in hours. To determine the actually eluted 99mTc activity, we must also take into account the elution efficiency, which is usually approximately 75-85%. This time dynamics of activity 99mTc during repeated elutions of the Mo-Tc generator is plotted in Fig.2.1.B (d), which we present here again for clarity :


Fig.2.1.B. Time dynamics of radioactivity in a mixture of two radionuclides.
a) In a mixture of two independent radionuclides
X , Y , each of them is converted according to its own half-life and the total activity of the preparation is given by the sum of both exponential functions.
b)
, c) In the decay series of two generically related radionuclides
X -> Y, the decay dynamics depends on the ratio of the half-lives of the primary parent radionuclide X and the daughter, further decaying radionuclide Y; depending on this relation lX and lY a transient or secular equilibrium of both radionuclides can then be established.
d) Specific radioactive dynamics of the radionuclide molybdenum-technetium generator during repeated elutions of the daughter
99mTc, resulting from the conversion of the parent 99Mo.

After elution, the 99mTc activity in the generator drops to almost zero, then rises and reaches a (local) maximum 23 hours after the previous elution, after which a radioactive equilibrium occurs and the 99mTc activity decreases exponentially with a half-life of 67 hours of 99Mo. After 23 hours from the last elution, the elution yield of 99mTc is the highest; the generator can of course be eluted as needed even in a shorter time, but with a lower yield of 99mTc.
   These elution cycles can be repeated many times until the activity of the parent radionuclide falls below the usable value; for a Mo-Tc generator with an initial activity of approx. 10-40
GBq it's about 7-15 days. The relatively long half-life of the parent radionuclide allows for long-term use of the generator, and the short half-life of the resulting daughter radionuclide ensures a low radiation exposure to the patient.
Note: In the past, have ben used sparadically also generators of the extraction type (by passing the methyl ethyl ketone through an aqueous solution of 99Mo, extracting pertechnetate 99Tc and separating it from the aqueous phase from the parent molybdenum), and sublimation type (using the difference between the volatility of molybdenum oxide and the resulting technetium oxide). Due to their excessive complexity and operational difficulties, they are no longer used, they have been pushed out by elution generators.
   Detailed decay scheme 99mTc is in Fig.4.8.2. Default metastable level 142keV isomerically passes first at the level of 140.5 keV, from where it emits primary gamma rays of energy 140.5 keV. With a very small proportion of 0.02%, there is a direct deexcitation to the ground state, in which energy of 142.7 keV is emitted. Photons of very soft gamma radiation of 2.17 kV are practically not observed, as they are almost 100% subject to internal conversion. The core of technetium 99Tc in the ground state (after the isomeric transition from 99mTc) is beta-radioactive and with a very long half-life of 200,000 years it slowly transforms into stable ruthenium 99Ru. In a very small percentage, there is a direct beta-conversion of 99mTc from a metastable level of 142keV to 99Ru (the ground state of the 99Tc nucleus is "bypassed") - mainly to excited levels of 322 and 90 keV 99Ru. Their deexcitation produces g- radiation with energies of 322, 232 and 90 keV, but a very small representation. At 99mTc radioactivity, soft characteristic X-ray with energies of »2-3keV (L-series) and »18-22keV (K-series) are also emitted, as well as a larger number of low-energy conversion and Auger electrons (approx. 4 electrons/1conversion), mostly with energies »1.6-3 keV, smaller amounts »120-140 keV.
   In the standard 99mTc scintillation spectrum (on a scintillation spectrometer or gamma camera) we observe only one significant photopeak of 140keV energy - Fig.4.8.3 on the left (on a semiconductor spectrometer we can also distinguish a weak line of 142.7keV) . Weak peaks from excited levels of 99Ru (arising from 99mTc by "bypass" 99Tc), especially 322keV, can be seen spectrometrically only after filtering out a strong over-radiating line 140keV with a layer of about 4-5mm lead - Fig.4.8.3 on the right in the passage "Radionuclide purity".


Fig.4.8.2. Energy levels, radioactive transformations and beta and gamma radiation
99m Tc.
Left: Formation of
99mTc by b- transformation of 99Mo. Right: Detailed decay scheme 99mTc.
The energies of the individual nuclear levels are counted in the left part of the figure from the ground state of 99Tc, while in the right part they are determined from the ground state of 99Ru.

The pertechnetate anions 99mTcO4- bind relatively easily to a number of biologically important substances (after possible previous reduction of pertechnetate, eg with tin ions). 99mTc is able to create chelates with functional groups of various organic substances and thus provide a wide range of radioactive preparations differing in their kinetics in the organism and uptake in individual organs.
Radioactively 99mTc-labeled radiopharmaceuticals 
This produces technetium- labeled radiopharmaceuticals that, after application to the body, are selectively taken up in certain target tissues or organs, which can then be imaged by a scintillation camera on the basis of external detection of the outgoing
g radiation. Technetium-labeled radiopharmaceuticals are widely used in planar and SPECT, static and dynamic scintigraphy of the kidneys, liver, lungs, heart, brain and other organs, as well as in tumor diagnosis. We will briefly mention some of the most commonly used 99mTc-radiopharmaceuticals :
  In some applications, the eluate alone in the chemical form of sodium pertechnetate Na
99mTcO4- will suffice. It is mainly scintigraphy of the thyroid gland (because technetium ions behave similarly to iodine ions), examination of Meckel's diverticulum, dynamic radiocardiography. In other applications, 99mTc atoms chemically bind to complex biochemical molecules.
  For dynamic renal scintigraphy
(§4.9.2., "Dynamic renal scintigraphy") is most frequently used 99mTc- MAG3 (merkapto acetyl triglycine) for diagnosing tubular function and renal drainage and DTPA acid (diethylene triamino penta acetid acid) for capture of glomerular filtration. For static renal scintigraphy, it is then DMSA (dimercaptosuccinate), which accumulates in the kidney in proportion to the function of the relevant sites and remains fixed there in the cortical zone in the cells of the proximal renal tubules for several hours.
   Iminodiacetic acid derivatives -
99mTc HIDA (.......) or EHIDA (.........) are used for dynamic liver scintigraphy (cholescintigraphy - §4.9.3 "Dynamic liver scintigraphy"), which they are taken up from the bloodstream by polygonal liver cells and further pass through intrahepatic and then excretory bile ducts into the duodenum.
   For dynamic radiocardiography examinations
(bolus radiocardiography and equilibrium gated ventriculography - §4.9.4 "Radionuclide ventriculography", "Dynamic radiocardiography") uses 99mTc- labeled erythrocytes (mostly labeled in vivo by Sn-pyrophosphate premedication), which remain in the bloodstream for the duration of the dynamic study.
    For scintigraphy of tissue perfusion and their viability are used
99mTc- isonitrile complexes, which in the form of lipophilic cations, passively penetrate the cell membrane, enter cells and bind there to cytosolic proteins and in the mitochondria of viable cells. The radiopharmaceutical accumulates, depending on blood circulation, in healthy viable cells, while in cells damaged (eg due to ischemia) or even dead and replaced by scar fibrous tissue, no accumulation occurs. The distribution of the radioindicator in the individual sites of the examined tissue is then proportional to the regional blood flow and the viability of the tissue cells. For scintigraphy myocardial perfusion is most commonly used 99mTc-MIBI (methoxyisobutyl-isonitrile) and 99mTc-Tetrofosmin (§4.9.4 "Scintigraphy myocardial perfusion"). The isonitrile radiopharmaceuticals are also used for non-specific cancer diagnosis - show increased accumulation in viable cells with a higher energy turnover (via mitochondria). For brain perfusion scintigraphy is used 99mTc-exametazime HMPAO (hexamethylpropyleneamine oxime) - §4.9.8 "Perfusion scintigraphy brain".
   For lung perfusion scintigraphy,
99mTc-labeled macroaggregates MAA of serum albumin are applied, the particles of which are trapped in the capillaries of the pulmonary circulation, in proportion to the blood supply to the individual parts of the lungs. For ventilatory lung scintigraphy, an aerosol of a suitable radiolabeled inert preparation (usually 99mTc-DTPA)  is inhaled; a better option is to inhale an inert radioactive gas (eg krypton 81mKr, see below).
   For skeletal scintigraphy, labeled phosphate complexes (pyrophosphates and polyphosphates) are used, which are osteotropic and bind to hydroxyapatite crystals; allow to view bone reconstruction. The most commonly used is
99mTc-MDP (methylene diphosphonate) - §4.9.7 "Skeletal scintigraphy".
Other radionuclides for g scintigraphy 
From other radionuclides for single photon
(planar and SPECT) scintigraphy can be briefly named for example :
Thallium
201 Tl (as chloride) for scintigraphy of myocardial perfusion, which as analog potassium enters myocyte cross the cell membrane and accumulates there in proportion to the blood flow at a given site of the heart muscle.
Galium
67 Ga - citrate for scintigraphy of tumors and inflammatory foci.
Indium
111 In is also used for a similar purpose. Also in the form 111In labeled antibodies for immunoscintigraphy, eg labeled somatostatin analog 111In-pentetreotide (OCTREOSCAN) for the diagnosis of neuroendocrine tumors - §3.6, section "Diagnosis of cancer", passage "Molecular gamma imaging".
  For radionuclide cisternography or perimyelography, intrathecal administration of
169 Yb -DTPA or more preferably 111In-DTPA is used.


Fig ... Decomposition scheme and gamma spectrum of indium
111In

For ventilation scintigraphy of the lungs was previously used gaseous xenon 133Xe (for complex dynamic scintigraphy), now radioactive gas 81mKr krypton, obtained from the generator 81Rb (T1/2 = 4.85 hours) (EC) ® 81mKr (T1/2 = 13s). A stream of air, guided guided by a tube through a container containing a layer of parent radionuclide 81Rb, carries away the released daughter 81mKr, which the patient inhales, and a scintillation camera uses external detection of radiation g to show the distribution of this 81mKr in the pulmonary alveoli - static ventilation scintigraphy of the lungs. For more complex pulmonary diagnostics, it is appropriate to combine perfusion and ventilatory scintigraphy.
   The principle of the 81Rb / 81mKr generator is in the left part of figure. The parent rubidium 81Rb is fixed in the solid phase in a small column, through which a stream of elution air is passed by means of a fan (air pump with adjustable power). By radioactive decay of rubidium-81, the continuously released daughter gas krypton 81mKr is entrained by the passing air and led to the respiratory mask, from which the patient inhales a mixture of air and radioactive 81mKr. One-way valves are included in the circuit of the breathing mask, and a mixing valve for outside air is also connected to ensure free breathing. Exhaled air is led to the extinction vessel (volume approx. 30 liters), from which, due to the very short half-life of 81mKr, practically non-radioactive air emerges.
   During this examination of pulmonary ventilation, inhaled air with a trace content of radioactive
81mKr enters the pulmonary alveoli, while the emitted radiation of 191keV gamma is scanned by a gamma camera. The scintigraphic image of the site of reduced activity shows areas of the lung with impaired ventilation, where krypton-81m, and thus no air, does not get (either at all or reduced) - see §4.9.5 "Lung scintigraphy (nuclear pneumology)".


Generator 81 Rb / 81m Kr.
Left: Principle of generator operation. Middle: One of the technical arrangements of the Rb-Kr generator. Right: Decay scheme 81Rb and 81mKr; in the black field is the scintillation spectrum of gamma radiation 81mKr.

Radionuclides and radiopharmaceuticals for PET
Of the more than 100 positron radionuclides, most are not suitable for PET imaging - due to too short or long half-lives, inappropriate radiochemical properties, low positron content and high intensity of unwanted electron and hard gamma radiation. Only a few
b+ -radionuclides are available for the medical use of positron emission tomography, but in the chemical form of a number of radiopharmaceuticals, enabling the study of various metabolic processes in the organism. For diagnostic imaging by PET, coincidence detection of g- photon pairs formed in the tissue during annihilation of positrons from a radioindicator with the electrons in tissue is used (described in detail above "Positron emission tomography"). The following positron radionuclides are mainly used :

Radioisotopes gallium 68Ga <-versus-> fluorine 18F
In current nuclear medicine, the radionuclides 18F (by far the most) and 68Ga are most often used for PET. What are the advantages and disadvantages of these radionuclides in terms of use in positron emission tomography?
    Fluorine-18 as a radionuclide has very good properties for PET: High yield of positrons (97%) with relatively low energy (max. 633keV) and thus short range in tissue (approx. 0.9mm) - excellent PET spatial resolution. A cyclotron can produce very high 18F activities of many gigaBq. Its half-life aprox. 2 hours enables transport to nuclear medicine workplaces located approx. 200-300 km from the production cyclotron. In addition to the basic widely used 18F-fluorodeoxyglucose FDG, a number of other radiotracers (mentioned below) are available. Until recently, however, it was not possible to effectively label some peptides and monoclonal antibodies with fluorine-18.
    Galium-68 has a slightly smaller proportion of positrons (89%). The higher energy of the emitted positrons (2900keV) leads to a greater range of the positrons in the tissue (approx. 4mm), which results in poorer spatial resolution - reduced ability to detect small and closely spaced lesions. 68Ga has a relatively short half-life of 68 minutes, which limits delayed PET imaging in case slower radiotracer pharmacokinetics. Transport to more distant PET workplaces is also problematic or impossible. However, the possibility to obtain 68Ga using a commercially available 68Ge/68Ga generator allows its application in PET workplaces without a cyclotron. However, it is laborious and expensive, and the amount of eluted 68Ga activity is relatively small (for about 2-4 patients).
    These prevailing disadvantages of gallium-68 have led to efforts to develop methods by which even those ligands that have so far been labeled with 68Ga (such as ligands for somatostatin neuroendocrine receptors and especially prostatic anti PSMA) could be labeled with fluorine 18F. Significant successes have been achieved especially with the prostate-specific tumor antigen PSMA, where the existing 68Ga-PSMA-617 can be replaced by e.g. 18F-PSMA-1007 and other derivatives, including the theranostic 18F-rhPSMA-7 (cf. §4.9, section "Combination of diagnostics and therapy - theragnostics", passage "Radiohybrid theranostic radiopharmaceuticals").

Fluoro-deoxy-glucose 18 FDG
By far the most commonly used radiopharmaceutical for PET is 2-deoxy-2-
18F-D-glucose, abbreviated as fluoro-deoxy-glucose (18FDG). FDG metabolism is somewhat different from normal glucose metabolism. Like ordinary glucose, FDG has an affinity for cells with increased metabolism (increased need for sugar - glucose), where it gets through the appropriate transport proteins and is subsequently phosphorylated. However, unlike true glucose, FDG is no longer metabolized and therefore accumulates in the cell. As a result, there is a markedly increased accumulation of FDG in the tumor cells, so that the tumor foci appear with high contrast to tissue and blood background. Oncological diagnostics therefore makes up more than 90% of all PET examinations (§3.6, section "Diagnosis of cancer"). 18FDG is also used to examine the myocardium, where myocardial viability can be assessed based on FDG consumption.
    Glucose it gets from the extracellular space into the cells by passive transport through transmembrane proteins - glucose transporters. Upon entering the cell, glucose is phosphorylated by gluokinase to glucose-6-phosphate (analogous to FDG-6-phosphate). Normal glucose can then be converted to glycogen or metabolized to water and carbon dioxide. However, this metabolism does not occur in FDG, so FDG is "trapped" and tends to accumulate in cells. For FDG, the only way is to be excreted back from the cell - through glucose-6-phosphatase. In cells containing low glucose-6-phosphatase, the concentration of FDG is proportional to glucose consumption. In contrast, in tissues that contain a lot of glucose-6-phosphatase, the accumulation of FDG is lower than that corresponding to glucose metabolism. Dephosphorylation by glucose-6-phosphatase is generally very slow, so that the concentration of FDG-6-phosphate in the tissues is kept stable for several hours.
   After radioactive decay,
18F produces non-radioactive oxygen 18O and FDG-6-phosphate produces ODG-6-phosphate, which then undergoes cellular glycolysis as normal glucose. Unmetabolized 18FDG is removed by glomerular filtration in the kidneys and is excreted in the urinary tract (with an excretion half-life of about 2 hours).
   
18FDG was first used in non-tumor diagnostics to visualize local glucose metabolism in the brain and myocardial glucose metabolism. However, it now has a major application in oncology as a radioindicator for imaging the increased metabolic activity of tumor tissues. Conversely, it is not suitable for the diagnosis of brain tumors. And it is also not suitable for the diagnosis of prostate cancer - due to the relatively slow metabolism of these tumor cells and the proximity of the bladder with a significantly higher content of FDG.
18F Sodium -fluoride (NaF) 
is used for PET scintigraphy of the skeleton. PET shows areas in the bones with osteoblastic and osteoclastic changes that may be related to tumor remodeling, but also to benign skeletal changes. The advantage of Na
18F is high (up to 50% of the applied activity) and rapid absorption in the bones, together with the rapid degradation of unbound radioindicator in the blood. This leads to the acquisition of contrast images in a short time (less than 1 hour after iv application).
   NaF is an ionic compound of Na+ and F- ions. After iv application, NaF is delivered to the bones and fluoride ions diffuse through the blood capillaries into the extracellular fluid. 18F-ions are exchanged for OH-ions of hydroxyapatite Ca10(PO4)6(OH)2 to form fluoroapatite. Subsequently, the installation 18F-ions into the crystalline structure of hydroxyapatite in bone occur. Increased uptake of 18F-fluoride occurs in malignant bone lesions due to increased blood supply, increased permeability of capillary walls, and faster bone remodeling. The advantage of Na 18F is that virtually all 18F-fluoride that is transported to the bones by the blood is trapped in them and, conversely, binding to serum proteins is minimal. This leads to the rapid degradation of the unbound preparation from the circulation and the acquisition of "pure" contrasting skeletal images in a short time.
Special radiopharmaceuticals for "molecular imaging" *)
With the development of organic chemistry, biochemistry and cell biology, some radiopharmaceuticals have been developed whose labeled molecules have affinity for very specific cell types or processes at the subcellular level. With the help of scintigraphy and a suitable radiopharmaceutical, it is possible to purposefully examine not only the function of a certain organ or tissue, but also to selectively investigate a certain type of metabolic or transport pathway, such as enzyme or receptor binding or antigen-antibody reactions. For this purpose, special radiopharmaceuticals (both for diagnostics and for therapy) have been developed and are still being developed, which are characterized by their effects at the molecular level. With a bit of exaggeration, these methods of local measurement and imaging of the physiological response are referred to as "in vivo biochemistry".
*) Name "molecular imaging"does not, of course, mean that we perhaps visualize the molecules themselves (unfortunately we can't do that...), but we visualize the distribution of radioindicators that is the result and reflection of specific biochemical reactions at the molecular level.
   For oncological diagnostics it is important imaging of viable and proliferating tumor cells and tissues - §3.6, section "
Diagnosis of cancer". In addition to the above-mentioned and most commonly used fluoro-deoxy-glucose 18FDG, there are some other tumor radiopharmaceuticals :
  
18 FLT ( 18F-3-fluoro-3-deoxy-thymidine)
is a radiolabeled form of a pirimidine nucleoside. It accumulates significantly in proliferating cells - it shows the activity of the enzyme thymidine kinase, which characterizes the intensity of cell division. Because a substantial increase in the rate of mitosis and cell proliferation is a hallmark of malignant tumor tissue,
18FLT functions as a tumor-specific PET radioindicator. It usually provides more contrasting images of proliferating tumor lesions than 18FDG. It is particularly suitable for monitoring the response of malignant tumors to therapy. Chemotherapy and radiotherapy often cause an inflammatory reaction in the tumor and around the tissue, which significantly increases the accumulation of 18FDG, making it very difficult to assess the regression or progression of the treated tumor; therefore, 18FDG is not a completely ideal radioindicator for the response of malignant tumors to treatment, 18FLT is more suitable.
   Thymidine is essential for replication in dividing cells. After passage of thymidine through the cell membrane, it is phosphorylated, which is catalyzed by the cytosolic isoenzyme thymidine kinase-1 (TK1), and subsequently incorporated into DNA (during the DNA synthesis phase of cell cycle). However, incorporating fluorine to the 3' position in thymidine prevents FLT from further incorporating it into DNA. FLT monophosphate is not incorporated into DNA and the cell membrane is impermeable to it - it is therefore metabolically "trapped" inside the cells. PET display
18FLT detects the enzymatic activity of TK1, tracing the recovery of nucleosides from degraded DNA. The uptake and accumulation of 18FLT thus corresponds to the rate of cell proliferation. 18FLT serves as an indicator of changes in tumor cell growth. 11C-thymidine was used for a similar purpose.
   18 F-fluorocholine (18 FCH)
is a fluorine-18-labeled analogue of choline, the basic building block of cell phospholipid membranes. It is used to visualize phospholipid metabolism in tumors. It shows increased uptake in tumors of the brain, prostate, breast, lung, esophagus.

   Choline is an important component of phospholipids in cell membranes. Choline is phosphorylated to phosphorylcholine by choline kinase inside cells and, after several other biosynthetic processes, is eventually incorporated into phospholipids.
18Flurocholine behaves in the same way. Cells with high metabolism also have increased choline uptake due to higher requirements for phospholipid synthesis in their cell membranes. 11C-choline is sometimes used for positron emission tomography, but its disadvantage is the short half-life of 11C (20 min.).
   18 F-fluciclovin is used for scintigraphic PET diagnosis of prostate tumors, especially in recurrent disease.
   Fluciclovin is an analog of the amino acid L-leucine, which is here labeled by
18F. It accumulates in the tumor via amino acid transporters. An increase in transmembrane amino acid transport occurs in the prostate due to increased metabolism of amino acids for energy and protein synthesis. Unlike the natural amino acids, fluciclovin it is not metabolized and accumulates in tumor cells - a positive PET image of the tumor.
   18 FET (18 F-O- (2-fluoroethyl) -L-tyrosine)
is an analog of the amino acid tyrosine to which an
18F- labeled ethyl group is attached via an oxygen atom. It shows the accumulation of amino acids in cells. It is suitable for imaging brain tumors glyoms, their extent, to detect recurrence after therapy and its differentiation from necrosis.
   Tyrosine is one of the building blocks of protein. Increased uptake into cells is due to the higher content of L-type amino acid transporters. However, the fluorinated FET analogue, unlike tyrosine, does not enter protein metabolism. Therefore,
18FET accumulates in cells and maps the increased amino acid consumption due to increased protein metabolism in tumor cells. 18FET (unlike 18FDG or 11C-MET) is not absorbed in macrophages and allows tumor tissue to be distinguished from inflammatory tissue.
   Other labeled amino acids are sometimes used to diagnose brain tumors:
11C-methyl-L-methionine (11C-MET), 3-123I-iodo- and methyl-L-thyrosine (123I-IMT), two other L-tyrosine analogues L-11C- tyrosine and 2- 18F-fluoro-L-tyrosine. All these substances have similar properties.
   18 F-FMISO ( 18 F-fluoromisonidazole) and 18 F-FETNIM ( 18F-fluoroerythronitroimidazole)
are radioindicators depicting cellular hypoxia, which is important for tumor angiogenesis and for planning radiotherapy (radiosensitivity, oxygen effect - see §3.6, section "
Physical and radiobiological factors of radiotherapy").
   Nitromidazoles passively diffuse through the cell membrane into the cytoplasm, where they are reduced by intracellular nitroreductases. The resulting nitro-radical R-NO can be further reduced to R-NH
2, which as a strong alkylating agent reacts with macromolecules - with DNA, RNA, with proteins. However, in cells with sufficient oxygen, the reduced nitromidazole is rapidly (re) oxidized and removed outside, further reactions no longer proceed, and the reaction products do not accumulate there. In cells with low oxygen content, however, the above sequence of reduction reactions occurs to form more stable covalent bonds with biomolecules whose rate is inversely proportional to the intrrecellular oxygen concentration. 18F-labeled nitromidazole derivatives are thus "trapped" in hypoxic tissue cells. This accumulation of 18FMISO occurs only in cells with active nitroreductases, so that the accumulation of the radiolabel occurs only in living hypoxic cells, not in necrotic ones.
   Another PET radiopharmaceutical for hypoxia imaging is
64Cu-ATSM (acetyl-methyl-thiosemicarbazone).
   18F-Florbetaben, 18F-Flutemetamol, 18F-Florbetapir
are
18F-labeled polyethylene glycol stilbene derivatives with high specific affinity for beta-amyloid plaques. It is used for scintigraphic visualization of amyloidosis especially in Alzheimer's disease in the brain, it can also be used in myocardial amyloidosis.
Immunoscintigraphy 
An important methodology for "molecular imaging" is immunoscintigraphy, based on the highly specific nature of antigen-antibody immunological reactions. The antibody, labeled with the appropriate radionuclide (
99mTc, 18F, 68Ga, 111In, 131I, 123I), selectively binds to the appropriate tumor marker (antigen) after administration, after which we can locate the relevant tumor by external detection of gamma radiation using a gamma camera. The required antibodies are either of human origin or are obtained from the serum of immunized animals by the method of so-called lymphocyte hybridization, which allows the preparation of a homogeneous so-called monoclonal antibody - §3.6, passage "Monoclonal antibodies". They are used mainly in tumor diagnosis, but also, for example, in the diagnosis of inflammatory foci using antigranulocyte monoclonal antibodies such as 99mTc-sulesomab and 99mTc-besilesomab (§4.9.6 "Scintigraphy of inflammatory foci").
Neuro-endocrine tumors 
For gamma imaging of neuro-endocrine tumors (including pancreatic), radioindicators that bind to somatostatin receptors are suitable. For example,
68Ga-DOTATOC is used (11C-5-hydroxy-tryptophan (5-HTP) was also tested). This diagnostics can also be used to assess (predict) subsequent radionuclide therapy using 90Y- or 177Lu-DOTA radiopharmaceuticals - teranostics (discussed in §4.9, section "Combination of diagnostics and therapy - teranostics").
  Recently, bombesin-based gastrin-releasing peptide (GRP) receptors, labeled with e.g.18F, have been tested for tumor dignostics. Bombesin *) is a peptide composed of 14 amino acids. It is formed in the small intestine and antrum, has a stimulating effect on the pancreatic and gastric mucosa. It is a potent antagonist of the neurotransmitter gastrin releasing peptide (GRP). It has been shown that it can be expressed by several cancer cell lines, where it can endocrine stimulate the growth of tumor cells via bombesin receptors on the membranes of these cells. Thus, bombesin may be a tumor marker for prostate, lung, gastric, neuroblastoma and others.
*) The somewhat bizarre name "bombesin" originated at the discovery of this substance, which was first isolated from the skin of a frog fire-bellied toad (Latin Bombina bombina).
Prostatic tumors 
Prostate- specific membrane antigen PSMA appears to be very promising for imaging and therapy of prostate cancer. Radiolabeled small molecules of PSMA inhibitors bind with high affinity to prostate tumor cells
(which highly express PSMA), allowing scintigraphic imaging of these lesions as well as their radionuclide therapy - depending on the radionuclide used. For scintigraphic imaging planar and SPECT can be used 99mTc-MIP-1404, for PET imaging of 18F-DCFBC or 68Ga-HBED-PSMA, recently 68Ga-PSMA-11. However, 18F-PSMA (18F-PSMA-1007) is best suited for PET scintigraphy of prostate tumors.
  
131I-MIP-1095 has been tried for targeted radionuclide therapy of the prostate, but 177Lu-J591 and more recently 177Lu- or 225Ac-PSMA-617 have proven to be the best so far.
   In addition to diagnostics, the possibilities of therapeutic use of monoclonal antibodies as carriers of suitable radionuclides beta or alpha with radiotherapeutic effect, or suitable chemotherapeutic agents, are also being developed. The use of monoclonal antibodies (non-radioisotope) in chemotherapy (biological treatment) of cancer is discussed in §3.6, section "
Therapy of cancer", in radionuclide therapy in the section "Radioimmunotherapy".
Labeled cytostatics 
Methods for radioactive labeling of various types of cytostatics have been developed for monitoring and prediction of chemotherapy of cancer (cytostatics are discussed in more detail in §3.6, section "Therapy of cancer"). The diagnostic application of such radiopharmaceuticals makes it possible to show where these cytostatics are taken up and to what extent they penetrate into tumor foci. In this way, the relevant cytostatics will be captured even during the chemotherapy itself - according to which the effectiveness of the treatment ("theranostics") can be inferred.
   One such radiolabeled cytostatic is 18F-paclitaxel (currently being tested in preclinical studies). For preliminary mapping of the distribution of some cytostatics, the 99mTc-MIBI, which has similar cell uptake kinetics to doxorubicin and cisplatin, can also be used.
Apoptotic radiopharmaceuticals 
A new interesting group are radiopharmaceuticals for imaging cellular apoptosis. These are organic molecules (either proteins or relatively small molecules) labeled with a suitable radionuclide (eg 99mTc, 18F), which have an affinity for cells that are in the early stage of apoptosis (programmed cell death - see §5.2 "Biological effects ionizing radiation", passage "Mechanisms of cell death"). Proteins bind to their surface (to phospholipids exposed on the surface of apoptotic cells), small molecules penetrate the cell membrane and accumulate in the cytoplasm. The result is selective accumulation of radioindicator in apoptotic cells and tissues. By gammagraphic imaging of the distribution of these radioindicators, we obtain positive images of those places where apoptosis occurs most intensively - whether due to irradiation, cytotoxic agents or ischemia.
    By molecular imaging of the distribution of cell apoptosis, we can monitor the very early response of cells and tissues to therapy (radiotherapy or chemotherapy), already at the beginning and during therapy - see §3.6, section "
Diagnosis of cancer". They can also be used to image ischemic foci in heart or cerebral infarction. Two types of such radiopharmaceuticals have been successfully tested in clinical practice for imaging apoptosis (the third is in the laboratory stage) :
- 99mTc-annexin V - is a protein that binds to phospholipids exsposed on the surface of cells undergoing apoptosis. Annexin V is obtained from the placenta and is labeled with technetium via hydrazino nicotinamide: 99mTc-6-hydrazinonicotin (HYNIC) -annexin V. For similar properties is tested 99mTc-Duramycin ;
- 18F-ML-10 [2- (5-Fluoro pentyl) -2-methyl malonic acid] - penetrates the depolarized cell membrane and accumulates in the cytoplasm of apoptotic cells;
- Peptide 18F-CP18 [pentapeptide containing triazole] - maps Caspase-3 activity, accumulates in apoptotic cells.

Radionuclides for therapy in nuclear medicine (methodological note)
Open radionuclide therapy is also organizationally integrated into the field of nuclear medicine (the main method of which is the scintigraphy discussed here). However, from our physical point of view, as well as from the point of view of the mechanism of action and purpose of use, we have included this radioinuclide therapy in §3.6 "Radiotherapy", part "Radioisotope therapy with open emitters". This is also where we find radionuclides used in nuclear medicine for therapeutic purposes.

Preparation of radiopharmaceuticals
As mentioned above, radiopharmaceuticals are composed of two basic constituents: a radionuclide emitting ionizing radiation and a carrier to which it is bound and which brings it to the required target in the body - to certain cells, target tissues and organs. For the preparation of radiopharmaceuticals - radioisotope labeling - three basic radiochemical methods are used :
× Isotope exchange reaction ,
wherein a certain stable isotope in the carrier compound is chemically replaced (exchanged, "displaced") by its radioactive isotope added to the reaction mixture. The resulting labeled substance has the same chemical and biological properties as the starting substance, because its molecules are chemically identical to the original molecules.
× Chemical synthesis ,
in which radioactive atoms are chemically incorporated into the appropriate site in the carrier molecule, most often by means of a coordination covalent bond, in such a way that the resulting complex compound has the desired properties. For this labeling, so-called chelates
(such as EDTA, DTPA) are often used, which by one part binds to the carrier molecule and their other part binds the radionuclide atom.
× Biochemical synthesis
uses enzymes and microorganisms. The radionuclide, added to the culture medium, enters metabolic processes in living microorganisms and then incorporates them into their respective metabolites.
   In terms of organization of preparation, we can divide radiopharmaceuticals into two groups :  
¨ Finished radiopharmaceuticals - mass production,
produced and radiolabeled in the manufacturer's radiochemical laboratory, delivered to nuclear medicine departments (appropriate activity and volume) and ready for direct application to patients. In this way, radiopharmaceuticals labeled with radionuclides with a longer half-life (> approx. 2 days) are supplied. These are, for example, radiopharmaceuticals labeled with iodine
131I and 123I, then 111In, 201Tl, 67Ga, 169Yb, recently also short-term 18F and others.
¨ Radiopharmaceuticals prepared at the workplace - individual preparation.
The required biochemical substance - carrier - is marked with the necessary radionuclide at the nuclear medicine workplace, in the laboratory of radiopharmaceuticals (preparation "magistraliter"). In this way, mainly radiopharmaceuticals labeled with short-lived radionuclides are prepared, mainly technetium
99mTc from a generator, sometimes also 18F. The actual synthesis was previously performed using basic chemicals, now the kits are used (kit = set of tools, building parts) - a compact set of non-radioactive ingredients supplied by the pharmaceutical manufacturer, to which only the solution of the radionuclide itself is added and the corresponding labeling reaction already takes place automatically.
   In nuclear medicine workplaces, the radiopharmaceutical laboratory deals with the preparation and filling of radiopharmaceuticals for their application. It is usually performed in special boxes or fume hoods equipped with air conditioning, ensuring laminar air flow. Current requirements for air cleanliness
(often exaggerated! - see below) *) lead to very complex and expensive air conditioning systems. New alternative solution, providing (without hood and air conditioning) sterility of radiopharmaceuticals and radiation protection of workers, are compact automatic devices for computer-controlled filling of radiopharmaceuticals, sometimes supplemented by the possibility of automatic application of the prepared solution of the radiopharmaceutical to the patient. They are mainly used in PET for 18FDG.


Preparation and filling (dosing) of radiopharmaceuticals at the workplace of nuclear medicine.
Left: Laminar hood for elution of Mo/Tc generator, preparation and dosages of radiopharmaceuticals at the Department of Nuclear Medicine, University Hospital Ostrava.
Right: Compact device for automatic computer-controlled dilution and dosage of radiopharmaceuticals.

*) Author's note - exaggerated requirements for the preparation of radiopharmaceuticals in nuclear medicine workplaces
From the point of view of my long-term work in the field of nuclear medicine, I would like to make a small critical comment on current standards and regulations for the preparation and filling (division, dossage) of radiopharmaceuticals in nuclear medicine workplaces. Until the 1990s, the actual laboratory radiochemical preparation of radiopharmaceuticals was carried out at workplaces,
in which the compounds prepared in the laboratory of radiopharmaceuticals by chemical procedures (in test tubes, beakers, penicillins) were labeled with the given radionuclide. This was done in conventional chemical fume hoods with possible lead shielding (sometimes with air extraction, other times not...), located in standard laboratory rooms. Following the principles of good laboratory practice, there have never been any problems with the sterility of the resulting radiopharmaceuticals ("nothing has ever happened to any patient "). Since the 1990s, kits (significantly facilitating preparation) and already done radiopharmaceuticals have been used more and more, that come sterile and are only filled and dosing into workplaces for administration to patients.
   And at this time, paradoxically, the requirements for the sterility of the environment began to appear and continue to tighten, in which these already significantly simpler manipulations are performed..!.. This leads to enormous investment and operating costs, which in my opinion virtually useless
(throwing hundreds of thousands even millions!)... The core of misunderstanding here is the confusion of minor "magistraliter" preparation in the workplace (or even just filling), to which officials try to mechanically transfer the demanding requirements, standards and regulations from the mass production of drugs in pharmaceutical plants (where these strict standards are, of course, justified). A doctor in an internal infirmary when drawing up a sterile injection preparation from an ampoule or penicillin into a syringe for i.v. administration, also does not work in an aseptic environment of class "A" (it would be nonsense)..!..
  Furthermore, narrow-minded standards and regulations regarding "registration" are a serious limiting factor in the introduction of new promising radiopharmaceuticals, already proven in research laboratories. This leads to the lagging behind of the field of nuclear medicine, at the expense of more advanced diagnostics and therapy of patients.
  
After all, we encounter similar bureaucratic approaches in the field of radiation protection (cf. §5.8, concluding note "Bureaucratic requirements of radiation protection").
Methods of administration of radiopharmaceuticals 
In terms of application form, three types of radiopharmaceuticals are used :
- Parenteral radiopharmaceuticals administered most often intravenously, sometimes subcutaneously or intralubally. They are mostly aqueous solutions, dispersions, colloids, suspensions. There are high demands on sterile and apyrogenicity.
- Oral radiopharmaceuticals can be in the form of solutions or solids. The most common are solutions or capsules of radioiodine given during thyroid therapy, or liquid or solid bites swallowed during examination of the esophagus and evacuation of the stomach.
- Inhaled radiopharmaceuticals are primarily radioactive gases (such as krypton 81mKr) or gaseous dispersions of labeled radiopharmaceuticals (eg 99mTc DTPA) produced in nebulizers, inhaled together with air during examination of pulmonary ventilation.
Quality and purity of radiopharmaceuticals 
The properties of the radiopharmaceutical used primarily affect scintigraphic diagnostics; unsuitable, poor quality or contaminated radioindicator can lead to inaccurate or erroneous diagnosis, ineffective radionuclide therapy, or it can also have undesirable side effects for the patient. From our physical and methodological point of view, the purity of the radiopharmaceutical is an important property, which can be divided into two categories :
v
Radionuclide purity 
Nuclear reactions, which produce their own radionuclides
(see §1.3 "Nuclear reactions and nuclear energy", part "Types of nuclear reactions" and §1.4, part "Production artificial radionuclides") used for labeling radiopharmaceuticals, usually take place in various ways and, in addition to the desired radionuclide, can lead to the formation of other radionuclides (the same element or another element). The amount of these radionuclide impurities depends on the target used, the type and energy of the irradiating particles and subsequently also on the method of separation and isolation of the given radionuclide.
There are three basic sources of radionuclide impurities (+ one special for radionuclide generators) :
1. Target material

can never be prepared in 100% "mononuclide" purity of the desired target nuclide. Traces of other isotopes of a given element, or even other elements, are always present. Nuclear reactions can then produce radionuclides other than the desired ones in the target from these impurities.
2. Different nuclear reactions - even with the same nuclide composition of the target, nuclear reactions can take place through different "channels" with different probabilities. When irradiated with neutrons, these are most often reactions (n, g), but they can also occur (n, p) or (n, d), etc., when proton irradiated, then reactions (p, g), (p, n), (p, d) and the like; it essentially depends on energy. Even in a completely pure target, a mixture of different radionuclides can be formed.
3. Radiochemical separation of a mixture of radionuclides formed by nuclear reactions during irradiation is a technologically difficult process, which may not succeed with 100% efficiency. Trace amounts of other radionuclides - radionuclide impurities - may thus be present in the final product.
4. For generator radionuclides, the radionuclide impurity can enter the desired daughter radionuclide in two ways :
¨ From the radionuclide impurities contained in the parent radionuclide.
¨ Traces of the parent radionuclide may also be released into the daughter radionuclide. E.g. in the Mo-Tc generator a small amount of parent 99Mo can penetrate into the 99mTc eluate, or in the Ge/Ga generator, a small amount of parent 68Ge may be released into the 68Ga daughter eluate.
   Radionuclide purity is the share of radioactivity of the required (declared) radionuclide in the total activity of the preparation. Usually, however, the opposite value is given - the content of radionuclide impurities - contaminants; it is mostly expressed as a percentage. The permissible content of radionuclide impurities is specified for each radioindicator in the relevant standard for its preparation
(eg for the 99mTc eluate, radionuclide impurities must not exceed 0.1%).
Measurement of radionuclide impurities
 
Accurate determination of the content of radionuclide impurities is performed by spectrometric measurement of radiation
g using a scintillation NaI(Tl) or semiconductor Ge(Li)/HPGe detector connected to a multichannel analyzer. It is not easy to measure the very low (trace) radioactivity of a contaminant in the background by many orders of magnitude higher activity of the basic radionuclide - the weak radiation of the contaminant is completely "over-radiated" by the radiation of the basic radionuclide. We have a chance to measure the radionuclide purity in basically two situations :
1. A high-energy contaminant
that emits gamma radiation with a significantly higher energy than the basic radionuclide. In this case, the method of filtration with a shielding absorbent insert can advantageously be used for the separate detection of the contaminant: place the vial with the examined preparation in a lead shield of suitable thickness (approx. 2-5 mm), which almost completely absorbs the intense low-energy radiation
g of the basic radionuclide, but transmits a considerable part of the weak but high-energy g -contaminant radiation.
2. Long-term contaminant
with a half-life several times longer than that of the basic radionuclide. If we measure such a preparation with an interval of 10 or more half-lives of the basic radionuclide, we obtain the activity or spectrum of the contaminant, which is no longer over-radiated by strong radiation of the basic radionuclide. This method has the disadvantage that it is an "ex post" measurement long after the preparation and use of the radiophamaceuticals. In some cases, however, there is no other option...
  A typical example of high energy contamination is the 99mTc eluate (Eg = 140keV), which may be contaminated with maternal 99Mo with a significant Eg = 740keV line. To shield strong 140keV radiation, we use a small lead container with a wall thickness of » 4-5mm; the transmitted 740keV radiation can already be measured with a scintillation detector, without the risk of being flooded by powerful primary radiation 99mTc, which is absorbed by Pb-shielding. For quantitative determination, it is of course necessary to have the detector pre-calibrated with a 99Mo standard for a given shielding and geometric configuration . The contaminant activity measured in this way is then divided by the total activity of 99mTc and we obtain a radionuclide impurity content of 99Mo.
  In Fig.4.8.3 in the left part there is a standard scintillation spectrum of
99mTc (a bottle with an activity of approx. 10kBq attached directly to the scintillation detector). Only a distinctive 140keV photopeak is displayed. In the right part of the picture there is a bottle with 1GBq of 99mTc eluate placed in a lead container with a wall thickness of 5mm. In addition to the residual (lead highly attenuated) peak 140keV on the spectrum, we see a peak of characteristic X-rays of lead around 80keV and of higher energies "climbed out" the two weaker peaks :
- 322keV comes from deexcitation of excited levels of ruthenium of 99Ru, to which a slight proportion of the beta -radioactivity breaks down the metastable level of 142keV 99mTc (see decay diagram in the previous Fig.4.8.2 on the right).
- 740keV comes from trace contamination of the eluate by mother molybdenum 99Mo (cf. previous Fig.4.8.2 on the left). It is from the intensity of this peak that the radionuclide purity of the 99mTc eluate is determined spectrometrically.
The same measurements on a semiconductor detector are at the bottom of the image.


Fig.4.8.3. Spectrometric measurement of the radionuclide purity of the
99mTc eluate (top - scintillation spectrum, bottom - semiconductor spectrum).
Left: Basic gamma radiation spectrum
99m Tc. Right: The spectrum of gamma radiation measured through the shielding layer of a 5mm lead container.

However, for a simplified measurement of the radionuclide purity of the 99mTc eluate, a conventional activity meter with an ionization chamber, equipped with a suitable shielding insert (and calibration), is used in most workplaces. However, due to the low sensitivity of these meters, we can determine up to radionuclide impurities of 99Mo exceeding hundreds of kBq (for Tc-eluates with an activity of tens of GBq, however, the sensitivity is sufficient to check compliance with the standard). It is sufficient to measure the radionuclide purity of the 99mTc eluate only for the 1st elution from the given generator, where the risk of potential contamination by 99Mo is the highest; if the result is satisfactory, it will almost certainly apply even more to further elutions.
v Radiochemical purity
Even chemical reactions by which radiopharmaceuticals are prepared by labeling with the required radionuclide do not proceed in 100% yield. Therefore, in the resulting preparations, in addition to the effective radioactive substance itself, there is always a small amount of unbound activity and possibly other compounds of radioactive substances that do not carry a diagnostic or therapeutic effect and may be disruptive or cause undesired radiation exposure in non-target tissues. Radiochemical purity is the share of the declared chemical compound of a given radionuclide in the total activity of the preparation.
  Chromatographic methods
(mostly on paper or on a thin layer) are most often used to control the radiochemical purity of prepared radiopharmaceuticals - §2.7, passage "RadioChromatography", sometimes also electrophoresis - see the same §2.7, passage "RadioElectrophoresis".
   Only a certain part of the molecules of a biological substance is marked by a radionuclide - we are talking about a radioactive substance with a carrier; the so-called carrier-free radioactive substance, where the radionuclide is contained in all molecules of the substance, is difficult to prepare and is used only for special purposes.

Application of radiopharmaceuticals
For their own use in nuclear medicine (diagnostic or therapeutic), radiopharmaceuticals are administered to the body, most often intravenously, or orally or by inhalation (as mentioned above). For each type of scintigraphic examination, a certain optimal amount of a given radiopharmaceutical is recomended, expressed in units of activity *) [MBq] of the radionuclide bound in the radiopharmaceutical - with normalization to body proportions, usually the patient's weight.
*) Radiopharmaceuticals cannot be dosed according to their weight [mg] - the weight of the "active substance", as is usual for drugs. The weights of applied radiopharmaceuticals are immeasurably small (often even on the verge of chemical provability). This is due to the high specific activity of the radionuclides used with a short half-life. The only way to dose radiopharmaceuticals is through the applied activity in [MBq]. The activity of the radiopharmaceutical for application is measured in a metrologically calibrated activity meter (§2.3, section "Well ionization activity meters"). Guideline values of the recommended applied activity for various types of radiopharmaceuticals are given in the table in §5.7 "Radiation exposure during radiation diagnosis and therapy", passage "Radiation dose to patients from radionuclide examinations".
   Other features of radiopharmaceuticals, such as sterility, apyrogenicity, content of excipients and other non-radioactive components, as well as details for the preparation of specific radiopharmaceuticals, are important for medical use, but their discussions, however, are beyond the scope of our physically and methodically conceived treatise ...


4.9. Clinical scintigraphic diagnostics in nuclear medicine
4.9.0. Common general principles of clinical scintigraphy 
General ideas of scintigraphic diagnostics were presented at the beginning of §4.1
(section "Role and definition of scintigraphy; nuclear medicine"). In the next text of chapter 4, the physical principles of gamma imaging, physical-electronic implementation of various scintigraphic methods and computer processing of scintigraphic data were discussed in detail. Here we will supplement this physical part with some specific clinical applications. But first we will make a few general remarks :
   The human organism
(as well as the organisms of all higher animals) is a very complex system both in its anatomical structure and, above all, in the diversity of its internal functions and metabolism. Imaging methods provide some valuable ways to "look" inside this complex system. Their basic output result is generally a brightness modulated image: the brightness of each element of the image is determined by the physical or chemical characteristics of the corresponding site in the tissue. In X-ray imaging, it is the absorption coefficient given by the density, thickness and composition of the tissue. For sonography, the acoustic impedance is displayed given the density, elasticity and dissipation viscosity of the tissue. Nuclear magnetic resonance displays the density of resonant nuclei (especially hydrogen) and relaxation parameters dependent on the binding of atoms in the tissue. Scintigraphy displays the distribution of a radioindicator, showing the movement of the radiopharmaceutical and its biochemical uptake, rearrangement and excretion due to local metabolic and functional processes (at the "molecular" level).
   Most imaging modalities - X-ray diagnostics, magnetic resonance, sonography - provide images of the anatomical structure of tissues and organs - their size and shape, placement, density inhomogeneities. Scintigraphic imaging (planar, SPECT, PET), on the other hand, has a fundamentally functional character. They do not show any "real-palpable" objects, no morphology, but capture the distribution - passage, accumulation, excretion - of specific radiolabeled substances. The degree of local accumulation of the radioindicator depends on the intensity of local metabolic and functional processes, which is reflected in the luminance modulation of the corresponding sites (pixels) of the scintigraphic image. If the radiopharmaceutical enters the examined tissue through the bloodstream, not only the function but also the degree of blood supply - perfusion of this tissue or organ can be assessed from the rate of uptake of the radioindicator
(§4.9.4, part "Scintigraphy of myocardial perfusion" and §4.9.8, part "Perfusion brain scintigraphy").
   Functional imaging of the distribution of the radioindicator in the relevant tissues, organs and lesions then serves primarily for diagnostic purposes, but can also be an important starting point for radiotherapy, in nuclear medicine for biologically targeted radionuclide therapy :


Fig.4.9.1 Scintigraphic diagnostics and radionuclide therapy in nuclear medicine

Static scintigraphy
Static scintigraphic images capturing radiolabel uptake are usually evaluated visually, but semi-quantitative assessments can also be performed using radiotracer accumulation ratios in relevant areas of interest (ROI) or tissue background. This results in certain relative numerical values - indices *). A series of scintigraphic images is evaluated both visually and quantitatively using computational algorithms
(see below "Radiopharmaceutical uptake" and "Dynamic scintigraphy").
*) Some software also allow comparison with databases of normal patients and determination of certain so-called scores - agreed quantitative parameters to facilitate decision-making between normal and pathological findings.
   Thus, radionuclide gammagraphy depicts functional and metabolic processes in tissues and organs, not their anatomical or morphological structure. However, scintigraphic images can provide certain information about morphology indirectly - by deriving from the display of the distribution of a radioindicator in the functional tissue of a certain organ or in tumor tissue. Non-functional tissue areas and, in general, areas where the radioidicist does not penetrate, are not displayed. It is often useful to supplement and combine ("merge") scintigraphic functional images with anatomical X-ray images (especially SPECT or PET with CT) - to perform a functional-anatomical correlation
(discussed above in §4.6 "Relationship scintigraphy and other imaging techniques", the "Mergers images of PET and SPECT with CT and NMRI"). This can lead to more accurate and comprehensive diagnosis.
   Disorders of function often precede disorders of anatomical structures - especially if the disorder is caused by altered molecular biochemical processes at the cellular level. Using radioisotope nuclear medicine techniques can therefore pathological changes often reveal earlier than other diagnostic procedures - even before structural changes are visible. Although scintigraphy has a smaller spatial resolution than X-ray or MRI, but thanks register individual photons
g it is very sensitive to subtle changes in the distribution of the radiopharmaceutical due to metabolic abnormalities.
Accumulation ( uptake ) of radiopharmaceuticals 
Rate of uptake - the accumulation of applied radiopharmaceuticals - in the examined tissues and organs is an important indicator of physiological or pathological function. This parameter uptake
(absorption) can be determined on the scintigraphic image based on the measured number of pulses in the marked region of interest (ROI) of the investigated structure, which is basically directly proportional to the accumulated activity. This number of photons emitted from the accumulated radioactivity lesion or organ towards the detector camera is, however, influenced by three factors :
- It is reduced by photons absorbed or scattered in the tissue located between the examined organ and the body surface
(for gamma 140keV it is about 50% per 6cm soft tissue).
- It can be increased by photons coming from a radiopharmaceutical collected in the surrounding tissues in front of and behing the organ under investigation - the body background.
- The most important factor: Radioactivity in the examined organ is determined by the applied activity and the share of accumulation of this total activity in the given organ
(this proportion corresponds to the functional status of the organ). The measured number of pulses is then directly proportional to the activity, acquisition time and sensitivity of the gamma camera.
   To determine the percentage of accumulation, we must first convert the activity applied to the patient in [MBq] to the corresponding number of pulses and acquisition time detected by the gamma camera. This can be done either by multiplying the camera sensitivity coefficient
(§4.5, passage "Sensitivity (detection efficiency) of the gamma camera") for a given radionuclide and collimator (if we have measured it in advance), or ad hoc by capturing an image of the applied activity (syringe) with the radiopharmaceutical (subtracting the activity of the "empty" syringe after administration and correcting for the half-life of the radiopharmaceutical and the acquisition times). Furthermore, it is desirable to make a correction for the absorption (attenuation) of gamma radiation in the tissue layer between the measured organ and the body surface (§4.3, passage "Adverse effects of SPECT and their correction", point "Absorption of gamma radiation").
  Determination of the percentage accumulation may be useful especially for assessing the accumulation capacity of the thyroid gland
(§4.9.1 "Thyrological diagnostics", especially before radioiodine treatment - § ..., passage "..."), relative renal function (§4.9 .., "Nephrological diagnostics") and accumulation of radiopharmaceutical in tumors during biologically targeted radionuclide therapy (§3.6, part "Radioisotope therapy"). In these images of tumor lesions (especially on tomographic PET images with 18FDG), a standardized value of accumulation SUV is often used to assess tumor viability (described in more detail above in §4.2, section "Quantification of positive lesions on gammagraphic images - SUV").
Specificity of scintigraphic diagnostics 
The specificity of scintigraphic methods may be lower or higher than for other modalities, depending on the mechanism of pharmacokinetics of the radioindicator used. E.g. in skeletal scintigraphy (§4.9.7 "Skeletal cintigraphy") bone metastases are seen earlier than on X-ray or MRI picture, but they cannot be distinguished from focal changes caused by other mechanisms of increased osteoblastic activity (inflammation, fractures) - high sensitivity but low specificity. In scintigraphic methods mapping the targeted and selective binding of a radiopharmaceutical to the cells of the examined tissue (tumor scintigraphy, perfusion scintigraphy of the myocardium or brain, receptor diagnostics), scintigraphic diagnostics can be highly specific.
Dynamic scintigraphy 
In addition to imaging and localizing structures in which a certain radiopharmaceutical accumulates physiologically or pathologically, it is sometimes important to assess the temporal dynamics of this accumulation or the passage of a radioindicator through the examined organs. This is the task of dynamic scintigraphy, capturing the time course of the distribution of the radioindicator - individual phases of the passage of the radiopharmaceutical through the examined organ - by means of a series of sequential images of the examined areas, scanned sequentially at selected time intervals *). The acquisition times of the images are chosen with respect to the speed of the studied process; they range from tenths of a second (for fast events such as cardiology) to tens of seconds or several minutes (kidneys, liver). The total scanning time is determined by the duration of the investigated process, ranging from one minute to 1-2 hours. The obtained series of images can be evaluated both visually (to observe the passage, accumulation and leakage of the radioindicator in various places), but above all quantitatively. In the pictures, we draw regions of interest (ROI) of significant structures, from which we construct curves of the time course of radiopharmaceutical distribution. We can then perform
mathematical analyzes of curves - for significant points and sections of curves various time intervals, ratios, integrals and other quantities are determined, appropriate functions are interpolated by least squares method, rate coefficients of radioactivity increase or decrease are calculated, curves are derived, integrated, filtration and deconvolution is performed, etc. - according to the used mathematical model of the investigated process. By this mathematical analysis of the time dependence curves of the indicator radioactivity in the relevant tissues and organs, we can obtain diagnostically important quantitative parameters of their function, both total and regional.
*) A special type of dynamic scintigraphy is phase scintigraphy of the cardiac cycle - ECG gated ventriculography. It is not created by simple sequential imaging, but by periodic recording and synchronous composing of a large number of consecutive images in different corresponding phases of the cardiac cycle - a detailed dynamic scintigraphy of one representative cardiac cycle is created (described in detail in §4.4 "Gated
phase scintigraphy ").
Functional parametric images 
When evaluating dynamic scintigraphy, we can obtain various quantitative parameters of the function, not only total, but also regional, local. In the extreme case, we can imagine that every image element - the pixel (i, j) of the image matrix - will be considered as a small elementary area of interest (microROI). From this microROI we can construct a curve of the time course of the radioindicator distribution at the corresponding place and mathematically process it using a certain model: eg interpolate an exponential or other suitable function and determine a certain diagnostically relevant dynamic parameter - eg velocity coefficient, slope gradient, increase or decrease half-life of radioactivity. The value of this locally calculated parameter is then stored in the same localized element (i, j) new image matrices, which we declare in the computer's memory. This is done for all pixels of the image of the examined area. By such processing of time curves from all microROI, ie from all pixels (i, j) of the image, we get a new artificial image, which no longer expresses the measured numbers of pulses on the scintigram, but shows the distribution of a diagnostically important parameter of the investigated function - functional parametric image, clearly mapping the local distribution of the function of the investigated organ. Thus, a parametric image in individual pixels, instead of the number of pulses, contains a number-value that characterizes a functional parameter.
   Functional parametric images are most often used in the evaluation of gated ventriculography - heart rate image, paradoxical image, Fourier images of phase and pulsation amplitude in individual places of the heart chamber (see below "
Gated ventriculography", or in more detail in "Radionuclide ventriculography" of the book "Ostnucline"). Special parametric images are provided by the so-called factor analysis (described in the works of M.Šámal and H.Trojanová ) .
Clearance  
Important global parameter that can be obtained by dynamic scintigraphy is the so-called clearance. It is the amount (volume) of blood or plasma in the body, that is "purified - cleaned" of a certain monitored substance per unit time. After a single intravenous administration of the appropriate radioindicator, we first monitor its accumulation and then its leakage from blood or plasma. The rate of decrease in plasma activity concentration *) depends on the elimination performance of the examined organs - kidneys or liver; with impaired function of the relevant elimination organs, the clearance value decreases. In dynamic scintigraphy, we construct a curve (histogram) from the area of interest outside the elimination organs, preferably from the precordium or lung area. This elimination curve represents the time course of the radioindicator concentration in the blood / plasma. We then evaluate this curve from the "blood pool" with compartmental analyzes - we interpolate bi- or multi-exponential functions using the least squares method. The elimination curve is basically created by the composition of two exponential functions. The first, faster exponential, is a manifestation of a relatively rapid balancing of the radioindicator concentration between the blood (intravascular) volume and the interstitial environment
(by filtering plasma from blood capillaries into tissue fluid to transfer nutrients and oxygen to cells) - dilution into the entire distribution area of the radiopharmaceutical. The second, more gradual component, is a reflection of the organ's own clearance by the elimination organ (kidney or liver). By interpolating the biexponential function, we mathematically separate both components, while the velocity coefficient in the later and slower exponential indicates the required value of plasma clearance. If the radioindicator used is excreted only in the organ under investigation, it is also the organ clearance value. The value obtained by this calculation is relative [sec.-1]; the actual (absolute) clearance in [ml./sec.] is obtained by multiplying the velocity coefficient by the value of the total distribution volume the indicator used in the organism. Calculation of clearance is described in the passage "Processing the blood-pool curve - clearance" in the book "OSTNUCLINE".
*) Plasma clearance was previously measured by sampling methods: after i.v. administration of the radioindicator, blood samples were taken at certain time intervals and their plasma activity was measured using scintillation detectors. Using simplified compartmental analysis, clearance values were calculated from them (formerly by graphical analysis on semi-logarithmic paper).

Transit functions, transit times
The elimination curve of the blood pool can be further combined with the organ curves and used for so-called deconvolution analysis, with the help of which we obtain the so-called transit functions of the passage of the radio-indicator through the elimination organs and their parts (e.g. parenchyma and hollow system of the kidneys, or liver parenchyma). It models a hypothetical situation, where we would apply the radio-indicator directly into the inlet vessel supplying the given organ and monitor the dynamics of the passage of this "bolus" of the radio-indicator through the examined organ.
The so-called transit times are also read on the transit curves, expressing the time (in seconds or minutes) for which the radio-indicator passes through the examined organ or its part. The minimum, mean and maximum transit times are evaluated.
   The mathematical description of deconvolution analysis, acquisition of transit functions, their analysis and interpretation is in the passage "
Deconvolution, transit functions", FIg.3.4.3 in §3.4 "Dynamic scintigraphy of the kidneys" of the book "OSTNUCLINE - Mathematical analysis and computer evaluation of functional scintigraphy".

Combination of diagnostics and therapy - theragnostics
Teragnostics
is generally a treatment strategy that purposefully combines diagnostics with therapy. New diagnostic imaging methods, especially molecular imaging in nuclear medicine, make it possible to integrate individual (personalized) diagnostics and targeted therapy
(or prevention) of serious diseases into a common field, for which the name teranostics or teragnostics was newly used (created by composing names: therapy + diagnostics => theragnostics). It's a kind of "diagnostic of the therapy". The analysis of the diagnostics and possibilities of therapy in an individual patient makes is possible to determine, whether the selected therapy will be effective, even before its initiation. And further assess the response to the performed therapy.
   In nuclear medicine, theranostics consists in the "pairing" of therapeutic and diagnostic radionuclides chelated to a suitable compound - carrier/vector - aimed at a specific clinical target (organ, tissue, group of cells). Due to their radioactivity, diagnostic radioisotopes emit gamma radiation
(either directly from the nucleus or during positron annihilation), which allows scintigraphic imaging of the target tissue. During their decay, therapeutic radioisotopes release charged energetic particles such as alpha, beta (possibly Auger electrons), which are capable of inactivating and killing cells in the target tissue due to their ionization effects. For such a theranostic pair ("matched pair of radionuclides"), the diagnostic counterpart can effectively predict the biodistribution of the therapeutic radionuclide.
   For teranostics in nuclear medicine is optimal when the atom of diagnostic gamma radionuclide
(such as 99mTc, 18F or 68Ga) has a similar chemical coordination of electrons in the valence shell as a therapeutic beta or alpha radionuclide( 90Y, 131I, 177Lu, or 225Ac). This allowing the same biochemical vector molecule to be used for scintigraphic diagnostics as well as for subsequent radioisotope therapy - just labeled with another radionuclide. The diagnostic drug is thus focused on the same molecular target as the subsequented therapeutic radiopharmaceutical, which allows to determine in advance the optimal applied activity and estimate the effectiveness of treatment --> theranostics :


Fig.4.9.2 Principle of teragnostics in scintigraphic diagnostics and biologically targeted radionuclide therapy in nuclear medicine.
Above: The same ligand-targeted biomolecule is labeled first with a diagnostic radionuclide (gamma or positron) and then with a therapeutic radionuclide (beta or alpha) using a suitable chelator. Bottom: Use of this created diagnostic radiopharmaceutical for scintigraphy, or a therapeutic radiopharmaceutical for biologically targeted radionuclide therapy.

Scintigraphy makes it possible to determine the concentrations of biologically active substances directly at the sites of their targeted action, which enables optimal and individual dosing, with the possibility of predicting effects and monitoring the results of therapy. We will first label the relevant biologically targeted substance with a diagnostic gamma-radionuclide, apply low activity and perform scintigraphic imaging. In case of successful uptake in target tissues (and sufficiently low unwanted uptake in healthy tissues - in critical organs), we label the same substance with therapeutic beta- or alpha- radionuclide and apply high activity to the patient (determined individually based on scintigraphy). It can be almost certainly assumed, that this therapy will be successful..!..
Theragnostic radionuclides 
Radionuclides suitable for theragnostics can be basically of three types :

1
. One radionuclide with mixed radiation beta- +gamma, beta- + beta+, alpha + gamma, or alpha + beta+ , which will be marked the relevant biologically targeted radiopharmaceutical. Gamma or positron emission allows scintigraphic imaging of planar/SPECT or PET. The emitted beta electrons or alpha particles then cause the radiobiological therapeutic effect of the desired destruction of pathological cells in the target tissue where the radiopharmaceutical has been taken up - theranostics. Such a radionuclide that is capable at the same time to enable diagnosis and therapy, we can (working) call it a "monotheranostic radionuclide".
   The best known example of such a "monotheranostic" radionuclide is the classical radioiodine
131 I , whose gamma radiation of energy 364keV allows scintigraphy (planar or SPECT), while beta- electrons can exert a therapeutic effect - with significantly higher applied activity. It has been used for decades in radioisotope diagnostics + thyroid therapy (hyperthyroidism and metastases of differentiated cancer, see below §4.9.1 ), although the name "theranostics" was not introduced at that time. More recently, the iodine-131-labeled monoclonal antibody tositumomab (Bexxar) has been used to treat lymphomas.
  In principle, some other "monotheranostic" mixed radiation radionuclides, such as lutetium
177 Lu, can be used for theranostic purposes. So far, the experimental teranostic radionuclide is terbium 149 Tb with mixed alpha-beta+ radiation: annihilation radiation from positrons+ can be used for PET diagnosis, emitted alpha particles cause a therapeutic effect (concomitant gamma radiation is not very suitable here for theranostics; but the main problem of 149-Tb is the complex conversion scheme with a number of secondary radionuclides - see 149 Tb ).
2. Two radioisotopes of the same element , one of which emits gamma photons or positrons beta+ for scintigraphic diagnostics, the second isotope emitting electrons beta- or alpha-particles for therapeutic effect. If we label the same biochemical with these two different isotopes, we get a diagnostic radiopharmaceutical for scintigraphy and a therapeutic radiopharmaceutical that will have identical biochemical properties to achieve successful theranostics.
   An example is iodine  
123 I for scintigraphy and iodine 131 I for therapy (used mainly in thyrology). So far, the experimental pairs of theranostic radionuclides are the positron isotope 64 Cu for imaging PET and the beta-isotope 67Cu for therapy, similarly 64 Sc/ 67 Sc, or the pair 86 Y for PET diagnostics and 90 Y for beta therapy.
3
. Two different radioisotopes of different elements, one of which emits gamma photons or beta+ positrons for scintigraphy, the second radionuclide is a beta- or alpha emitter for therapy. Each of these radionuclides is chelated to the same biochemical substance to give the appropriate diagnostic and therapeutic radiopharmaceutical. Their identical biochemical properties are no longer 100% ensured here, they can be influenced by various chelators and complex chemical bonds - they need to be carefully verified... In the positive case, the theranostic approach will also be successful. A new interesting method of 100% ensuring identical pharmacokinetics of diagnostic and therapeutic radiopharmaceuticals are the radiohybrid theranostic radiopharmaceuticals described below.
  Theragnostic method of is performed, so far mostly experimentally, using several radionuclides, for a number of monoclonal antibodies with the help of various chelators. For scintigraphic imaging of planar/SPECT/PET, the radionuclides 99m-Tc, 111-In, 18-F, 68-Ga, 89-Zr, 124-I are used, which are subsequently combined with the beta-radionuclides 90-Y, 177-Lu, or with alpha-radionuclides 212-Bi, 227-Th, 225-Ac - in each case the same monoclonal antibody. Recently, for example, the combination (68Ga/177Lu) -PSMA J591 or (68Ga/225Ac) -PSMA-617 at metastatic prostate ca appears to be promising.
Radiohybrid theranostic radiopharmaceuticals 
A newly developed interesting "trick" is to bind two chelators to a targeted ligand molecule simultaneously (hybridly) with two required atoms, not yet radioactive. For example, natural fluorine
19F and natural lutetium natLu (consisting of 97.4% 175Lu and 2.6% 176 Lu). If we then add to such a preparation the appropriate radioactive isotope - either 18F or 177Lu, it is labeled by the mechanism of isotope exchange with one or the other radionuclide. As needed, either 18F for diagnosis (lutetium remains inactive) or 177Lu for therapy (here conversely the fluorine remains inactive). It also automatically ensures the identical pharmacokinetics of the substance labeled with the diagnostic and therapeutic radionuclide in the sense of point 3. above, as the two molecules are atomically- chemically identical, differing only isotopically. It is an ideal feature for theranostics. It is currently being tested on the PSMA 18F/177Lu .
        
Fig.4.9.3. Principle of radiohybrid theranostics. Above: Ligand vector biomolecule with two bound chelators with non-radioactive atoms (here natural fluorine and lutetium). Middle: Binding of radioactive fluorine atoms
18F or lutetium 177Lu by isotope exchange with inactive atoms. Bottom: The use of thus created diagnostic radiopharmaceutical for scintigraphy (PET) or a therapeutic radiopharmaceutical for biologically targeted radionuclide therapy.
   This approach is currently being tested for PSMA, a radiohybrid rhPSMA-7 has been developed, primarily for 18F/177Lu labeling, using a silicon fluoride acceptor for efficient isotopic exchange of inactive 19F fluorine for beta+ radioactive 18F fluorine.

Radionuclide examinations in nuclear medicine
In nuclear medicine, a number of methods have been developed for radionuclide examinations of various tissues and organs in order to determine their normal or pathological states.

   In the beginning of the field of nuclear medicine (60s-70s of the 20th century), sample methods were often used  - blood or plasma samples taken from patients after application of a radioindicator were measured using radiometers (mostly cavity scintillation detectors). From the measured activity of these samples (in relation to the applied activity), the parameters of function - clearance, distribution volumes - of the respective radiopharmaceuticals in the monitored organs, or blood circulation were determined using mostly empirical methods, dilution principles, etc. Or, the concentration of the radiotracer in the body was measured using scintillation probes aimed at organs of interest (eg kidneys, heart). During the 80s-90s and the first decade of the 21st century, these methods were gradually abandoned. These were often "blindly" methods; from today's point of view, they provided less accurate and less reliable results with greater laboriousness, with the possibility of significant individual errors; so it is good that they are already abandoned...
   Now a simple "equation" applies, that :
              
  Current (and future) nuclear medicine = scintigraphy + radionuclide therapy with open radionuclides .
   The preparation of the patient before the scintigraphic examination depends on the examined process, the radiopharmaceutical used, the state of health and the patient's previous medication. Above all, before the examination, with a certain time interval, it is necessary to discontinue such drugs that would adversely affect the biodistribution of the radioindicator used, or the function of the examined organ
(for thyroid examination they are iodine preparations, for cardiac nitrates, beta-blockers, diuretics, cardiotonics). Prior to non-thyrological examinations, Chlorigen is often given to block the thyroid gland. Before to the actual scintigraphic examination, an appropriate radiopharmaceutical must be prepared at the nuclear medicine workspace (§4.8 "Radionuclides and radiopharmaceuticals for scintigraphy"), for i.v. application filled in a syringe with a specific value of activity optimized individually for each patient (usually based on body weight) *). The activity of the radiopharmaceutical for application is measured in a metrologically calibrated activity meter (§2.3, section "Well ionization activity meters").
*) Guideline values of the recommended applied activity for different types of radiopharmaceuticals are given in the table in §5.7 "Radiation exposure during radiation diagnosis and therapy", section "Radiation exposure of patients from radionuclide examinations".
   The time course of scintigraphic examination depends mainly on whether it is static or dynamic gammagraphy, what is the course of the examined process in the organism and how fast is the pharmacokinetics of radioindicator used. In static scintigraphy, the application of the radiopharmaceutical is usually performed off-camera (in the application room), the volume and speed of the application do not matter, the actual imaging is taken with a certain time interval, it is necessary to wait until the radio indicator is sufficiently taken up and accumulates in the required tissues - it can even be in 2 hours
(e.g for skeletal scintigraphy or for PET imaging of 18FDG accumulation in tumor lesions).
   In dynamic scintigraphy, the radioindicator is applied directly below the scintillation camera, the field of view of which is set to the patient's examination area, whereas dynamic imaging starting immediately with the application. For dynamic scintigraphy of fast processes (such as blood flow through the atria and ventricles during angiocardiography, or monitoring the dynamics of the perfusion phase in the brain or kidneys) it is necessary to perform a so-called bolus application: rapid application of a radioindicator with high activity in a small volume of about 0.5 ml. - compact bolus, with immediate start of dynamic scintigraphy with a sufficiently high frame frequency (one or more frames/sec.).

Methods of clinical scintigraphic diagnostics
Here we will briefly describe some more important methods of scintigraphic diagnostics. In the introduction to the individual areas of scintigraphic diagnostics, we will first present a brief outline of the structure and biological function of the examined tissue or organ and its most common pathologies, based on which we will analyze methodological approaches to the diagnosis of relevant disorders and diseases. For each specific method, we will state its medical purpose, the radiopharmaceuticals used, procedure of performing the examination, and finally its processing and evaluation, with examples of scintigrams and the results of normal and pathological
(all listed scintigraphic images were acquired and evaluated at the Clinic of Nuclear Medicine University Hospital Ostrava). Since there are a number of scintigraphic methods in nuclear medicine, we have divided this topic into several numbered subchapters according to the investigated organs, systems or separate issues :

4.9.1. Thyreological radioisotope diagnostics 4.9.2. Nephrological radionuclide diagnostics
4.9.3. Diagnostics of the gastrointestinal tract
         - liver and bile ducts, pancreas, esophagus and stomach
4.9.4. Nuclear cardiology
4.9.5. Pulmonary scintigraphy (nuclear pneumology) 4.9.6. Oncological radionuclide diagnostics
4.9.7. Skeleton scintigraphy 4.9.8. Scintigraphic diagnostics in neurology - CNS
4.9.9 Radionuclide phlebography and lymphoscintigraphy; Sentinel nodes. Labeled blood elements. 4.9.10. Less significant and rarely used radionuclide examination methods

4.9.1. Thyreological radionuclide diagnostics
The thyroid gland is located in the front part of the neck and, despite its small dimensions (approx. 5 x 7 cm), it is a relatively important organ, intervening in a number of processes in the whole organism. Thyroid function is closely linked to iodine metabolism in the body. Sodium iodide NaI penetrates thyroid cells by transport via the Na/I symporter (which is a 37 kDa transmembrane glycoprotein) - the "iodine pump". Within thyroid cells, iodine binds in the thyroglobulin molecule to form monoiodine- and diiodine-tyrosine. Their combination then produces thyroid hormones - triiodothyronine (T3) and tetraiodothyronine (thyroxine T4). Therefore, radioiodine is also efficiently taken up by the thyroid gland. Thyroid hormones enter cells in the body and are involved in regulating a number of metabolic processes in the body. They affect the transport conditions on cell membranes for the entry of sugars and amino acids into cells. T3 binds to the corresponding T3 receptors on the surface of mitochondria and thus regulates intracellular metabolism. It also binds to T3-responsive domains in nuclear DNA and initiates mRNA production for proteosynthesis in cells.
Thyroid pathology 
The most common functional pathologies of the thyroid gland are :

-
hyperthyroidism - increased thyroid function with excessive production of hormones (T3, T4).
-
hypothyroidism - decreased thyroid function.
- functional autonomy - independence of the function of certain areas in the thyroid gland on regulatory mechanisms.
   A common morphological disorder is an enlarged thyroid gland or goiter, which can be diffuse or nodular. Depending on the function, the goiter may be eufunctional, hyperfunctional or hypofunctional. Nodes - areas of increased density, occur quite often in thyroid tissue. And not only one node - unifocal, but also more nodes - multifocal disability. In terms of function, they may have the same function as the surrounding tissue, or they may be hyper- or hypofunctional.
   The most serious thyroid disability is its tumorous disease - thyroid cancer. From a histological point of view, there are 3 basic types of thyroid tumors :
- Differentiated adenocarcinoma, which is further divided into follicular, papillary and mixed. Follicular carcinomas (15-30% of all thyroid malignancies) are mostly unifocal and spread mainly through the bloodstream. Papillary and mixed cancers are the most common (30-70%), they are mostly multifocal and metastasize mainly through the lymphatic system. Differentiated thyroid carcinomas retain iodine accumulation and are therefore successfully treatable with 131I radioiodine (§3.6, section "Radioisotope therapy", passage "") .
- Medullary carcinoma (approximately 5-10% of the incidence) is based on parafollicular C-cells, it spreads mainly hematologically and its treatment is more difficult than in differentiated ones. It often does not respond to radioiodine...
- Undifferentiated (anaplastic) carcinoma (approximately 10% of the incidence) originates from follicular cells, metastasized by blood and lymphatic routes. It tends to be quite aggressive with invasion of surrounding tissues and the formation of more distant metastases. Its treatment is difficult and usually unsuccessful (it does not respond to radioiodine).
   From a general point of view, the issue of cancer is discussed in more detail in §3.6 "
Radiotherapy", radionuclide therapy, especially in the section "Radioisotope therapy", not only cancer, but also, for example, hyperthyroidism.
   Radioisotope diagnosis of the thyroid gland is the oldest method of nuclear medicine (first tested in 1938). This is due to the strong ability of the thyroid gland to accumulate iodine - and thus even the radioiodine, the radioactive isotopes. Previously, only simple accumulation tests were performed with 131I radioiodine, the amount of which in the thyroid gland was measured with a single gamma-probe; it was determined what percentage of the applied amount of radioiodine is taken up in the thyroid gland. Later, scintigraphic methods were introduced.
The thyroid accumulation test 
is now performed only before radionuclide radioiodine therapy to determine the applied activity. The patient is given about 0.5-1 MBq of radioiodine, orally in the form of a solution of sodium radioiodide. After its absorption from the GIT, iodine ions are taken up by the functional tissue of the thyroid gland
(or even by metastases of differentiated thyroid gland). After 6 or 24 hours, the captured activity in the thyroid gland is measured either by a simple collimated radiometric probe (see Fig.2.4.3 b in §2.4, passage "Scintillation probe") or by a gamma camera (with marking and quantification of ROI on the thyroid image) and compared with the activity of the administered radiopharmaceutical - the result is the percentage of radioiodine taken up *). Normal values are about 5-15% in 6 hours and 10-30% in 24 hours. This measurement can also be performed repeatedly over several days to determine the dynamics of gradual leakage (clearance) of radioiodine from thyroid.
*) The results may be skewed by some drugs containing iodine, which saturates the uptake mechanisms and reduces the accumulation of radioiodine - these must be discontinued !
   Measurement of radioiodine accumulation in the thyroid gland, as well as its clearance - effective half-life, is important for individual determination of the required applied activity of radioiodine to achieve optimal therapeutic effect in hyperthyroidism and autonomic adenomas - see §3.6, section "Therapy of thyroid gland with radioiodine 131 I", section "Individually applied activity - Marinelli equation".
Thyroid scintigraphy 
Purpose:
To show the distribution of functional thyroid tissue in the primary diagnosis, its location, shape and size and reveal possible anomalies - areas of increased or decreased function in thyroid tissue, finding ectopic thyroid tissue. Furthermore, showing the functional properties of palpable nodes and functional autonomy. In combination with laboratory determination of T3, T4, TSH levels, assessment of hyper- or hypothyroidism. In thyroid cancer therapy, scintigraphy is used to demonstrate residual accumulating tissue and to detect distant accumulating metastases of differentiated thyroid cancer.
Radiopharmaceuticals: 
The basic method consists in the oral administration of radioiodine
131 I in the form of sodium iodide Na 131I. Due to the higher radiation exposure (radiation b which is to diagnostic unusable), however, 131I is no longer used for primary diagnosis, it is replaced by an isotope of iodine 123 I-sodium , or 99 m Tc pertechnetate, which also taken up and accumulated in cells similarl to iodide, but unlike there from does not bind to thyroglobulin and does not enter into other metabolic reactions. 131I (application approx. 10MBq) is used only in patients with proven thyroidopathy before radioiodine therapy.
   To finding less differentiated tumor tissue and metastases are also used 
99mTc-MIBI and 99mTc-Tetrofosmin. It to image medullary carcinoma (which contains somatostatin receptors) is further used 111In-pentetreoid, also 123,131I-MIBG.
Execution : 
For scintigraphy of the thyroid gland, as a small organ, we use a high-resolution collimator using ZOOM, or a pinhole collimator (which increases the projection of the thyroid gland over a larger usable area of the camera's scintillation detector). After iv application of
99mTc-pertechnetate (approx. 100-200 MBq), a planar image is captured in the AP projection after about 30 minutes. For better morphological orientation, it is advisable to take a picture with marking using a point source - "pointer".
Evaluation :
The normal image of the thyroid gland has a like "butterfly" almost symmetrical shape, with an approximately homogeneous distribution of the radiopharmaceutical in both lobes of the parenchyma. The pathological picture shows an inhomogeneous distribution with "cold" nodes of reduced function, or "hot" nodes of increased thyroid tissue function. Possible functional autonomy of hot nodes can be determined by repeated suppression scintigraphy of the thyroid gland after several days of administration of triiodothyronine. In functional autonomy, the accumulation of the radiopharmaceutical in the "hot" deposit does not change, while in the others parenchyma (paranodular tissue), due to hormonal suppression, the accumulation decreases significantly or disappears (§3.9 "Quantitative thyroid scintigraphy" in OSTNUCLINE).

Normal
thyroid scintigram
Whole-body scintigraphy
after radioiodine therapy of the thyroid gland .

Multiple accumulating deposits in both lung wings - metastases of thyroid cancer - appeared.

Under favorable circumstances, these malignant foci can be successfully eradicated by radioiodine therapy. 
Hyperfunctional node in the
right lobe of gland
Nodular goiter with unfunctional
node on the left
  Examples of typical images of thyroid scintigraphy
(the pictures were taken by MD. V.Dedek, PhD., KNM Ostrava)
 

When searching for functional metastases of differentiated thyroid carcinomas, it is appropriate to use whole-body scintigraphy with radioiodine (application approx. 100-200 MBq). Thyrological diagnostics in nuclear medicine can then be followed by radionuclide therapy of the thyroid gland - treatment of hyperthyroidism, autonomic adenoma, thyroid carcinoma (see §3.6 "Radiotherapy", section "Radioisotope therapy with open emitters"). After therapy is followed by control scintigraphy of the thyroid gland or metastases at certain time intervals.
Parathyroid scintigraphy (parathyroid glands) 
The parathyroid glands are 2 pairs of small formations located on the back of both lobes of the thyroid gland. They are glands producing the parathyroid hormone, that affect the calcium content - releases it from the bones into the blood.
Purpose: Using scintigraphy, we try to show the hyperfunctional parenchyma of enlarged parathyroid glands (usually their adenoma), which by their increased production of parathyroid hormone, adversely affect the turnover of calcium in the body.
Radiopharmaceuticals: As there are no radioindicators that are selectively taken up in the parathyroid glands, the cationic complexes
99mTc-MIBI and 99mTc-Tetrofosmin are used, but they also accumulate in the thyroid parenchyma.
Execution and evaluation: Displaying small parathyroid glands against the background of much larger thyroid tissue is not easy. We can help in two ways :
- Two-phase scintigraphy using faster leaching of
99mTc-MIBI from thyroid tissue, than from the parathyroid gland affected by enlargement or adenoma. In about 15-30 minutes after iv application of the radiopharmaceutical (approx. 700-800MBq) we take the first image, the next in 2-3 hours.
- Subtraction scintigraphy performing computer images subtraction. The image of the thyroid gland taken after the application of
99mTc-pertechnetate alone (approx. 200MBq) is subtracted from the "summation" image [of the thyroid gland + parathyroid glands], taken after the subsequent application of 99mTc-MIBI or 99mTc-Tetrofosmin. Both images must be captured under identical conditions, without changing the position. After subtracting the image of the thyroid gland tissue, in the remaining "summation" image, the image of the parathyroid gland itself remains.

4.9.2. Nephrolological radionuclide diagnostics
The urinary excretory system, formed by the system [kidneys - ureters - bladder - urethra], is collectively called the uropoietic system.
The kidneys *) are mainly serves to filter blood (which is supplied to the kidneys by the renal arteries), which removes metabolic products and other unnecessary or harmful substances from the body, which then flow out as urine out of the body. Waste products - catabolites - of nitrogen metabolism (urea, creatinine), acid catabolites, water and electrolytes are thus removed, thus maintaining a stable internal environment. The kidneys also have a regulatory function- ensure homeostasis of the organism - water, salts, minerals, acid-base balance, participate in maintaining blood pressure. The basic building block and functional unit of the renal parenchyma is the nephron. The human kidney contains about 800,000 to 1.5 million nephrons.
*) The kidneys are called ren in Latin, and nephros in Greek - hence the synonym for renography = nephrography in the examination methods.
    The nephron begins with a small ball of capillaries called the glomerulus, where the branching renal artery brings blood. In the glomeruli, the basic clearance function of the kidneys takes place - glomerular filtration, which is a process of ultrafiltration blood plasma under pressure across the glomerular membrane. The microporous structure of the glomerular wall (which prevents the flow of plasma proteins larger than about 100 kDa) and the electrostatic barrier of the glomerular membrane (negative charge of polyanionic macromolecules of the membrane and negative charge of most plasma proteins prevent the transfer of even smaller proteins with molecular weight above about 60). The glomerular ultrafiltrate is primary urine, which is essentially plasma without cells and large molecules of plasma proteins; more than 150 liters of this liquid are made in kidneys a day. In addition to waste metabolic products, it contains a number of substances and nutrients (such as glucose) that should remain in the body.
    The glomerular filtrate enters a hollow canal of the nephron, called the tubule. Here, tubular resorption takes place, during which part of the substances from the glomerular filtration is reabsorbed and returned to the blood, leaving through the vascular bed around the tubules, through a drainage vessel from the kidney. Most water, glucose, amino acids, minerals return to the blood by resorption. This maintains homeostasis - the balance of water and salt in the body. In addition to resorption, tubular secretion also occurs here - tubule cells actively take up some substances from the blood (eg creatinine) and transport them to the tubular cavities, ie to the urine. The total amount of substance excreted in the urine is given by the sum of: (glomerular filtration) - (tubular resorption) + (tubular secretion).
  After passing through the tubular system, definitive secondary urine is formed, which consists of water with dissolved urea, sodium chloride and a small amount of other substances; about 1.5-2 liters per day is excreted. The tubules open into thicker collecting ducts, funnel-shaped calyxes and finally into the renal pelvis (pelvis renalis) of the hollow kidney system, from where urine flows through the ureters into the bladder. From there, after releasing the sphincter, it flows out of the body through the urethra.
Pathologies od the kidneys and urinary tract  
The kidneys are relatively often affected by inflammatory and infectious diseases. Pyelonephritis is a bacterial purulent inflammation of the kidneys (renal pelvis - pyelos as well as parenchyma), which can be acute or chronic and can lead to deterioration in renal function if repeated or prolonged. Glomerulonephritis is an inflammatory disease affecting mainly the glomeruli in the kidneys, which can occur after infections (especially streptococcal), autoimmune processes and other causes. Some kidney diseases can lead to deterioration of renal function, which can be irreversible (nephron loss), in extreme cases can result in kidney failure...
  Very common kidney and urinary tract involvement is lithiasis (urolithiasis) - "kidney stones", formed by the accumulation and increased concentration of mineral salts that crystallize in the urinary tract, especially in the pelvis of the kidneys or in the bladder. The most common are stones from calcium oxalate or uric acid. They can grow to various sizes, from small particles of "sand" to larger stones (> 1 cm), which can block the outflow of urine from the kidney. This ureteral obstruction causes congestion in the hollow system of the kidney, which also burdens the parenchyma (which must filter against pressure); with prolonged obstruction, renal function is irreversibly impaired.
    When renal function is impaired, glomerular filtration is reduced and thus waste products are retained in the body and tubular resorption is reduced, and the absorption of electrolytes and water is impaired. This also affects blood pressure and can lead to disorders of acid-base balance, or even hematopoiesis.
    
The renal parenchyma can be affected by cystic (often polycystic) disease. Renal tumors, such as Grawitz's tumor, are relatively uncommon. .....
    The kidneys and their excretory functions have already become a suitable object for radioisotope diagnostics in the beginning of the field of nuclear medicine.
Until the 1970s and 1980s, radioisope renography was one of the most frequent nuclear medical examinations. After application of the nephrotropic radioindicator 131I-hippuran, the course of radioactivity in the kidneys was recorded with two collimated scintillation renographic probes (see Fig.2.4.3b in §2.4, passage "Scintillation probe"), attached to the patient's back in the kidney location. The electrical signal from the detectors was fed to a double registration recorder, the pen of which plotted the so-called nephrographic curves on paper. From the shapes of the nephrographic curves it was possible to deduce various pathological conditions and disorders of renal function, as well as urine outflow from the kidneys. Semi-quantitative analysis of nephrographic curves was sometimes performed. However, it was only an approximate "blindly" examination, without the possibility of regional assessment. Replacing isotope nephrography with dynamic scintigraphy has significantly refined diagnostics and provided much more comprehensive informations :
Dynamic renal scintigraphy 
Purpose :

It is the most important method of nuclear nephrology. It is used for a comprehensive assessment and quantitative analysis of perfusion and excretory function of the kidneys (and their parts), clearance and drainage - the dynamics of urine outflow from the kidneys. It can also provide certain information about the morphology of the kidneys, which, however, is derived from the display of the radioindicator distribution in the functional tissue of the kidneys (parenchyma) and from the outflow or accumulation of the radiotracer in the hollow system.

Radiopharmaceuticals:
- 99mTc-DTPA (diethylenetriaminepentaacetic acid), which is excreted by passive ultrafiltration in the glomeruli (not resorbed in the tubules). In addition to functional imaging, it is also suitable for the determination of glomerular filtration by plasma clearance.
-
 99mTc-MAG3 (mercaptoacetyltriglycine), which binds to transport plasma proteins and is excreted by tubular secretion (in glomeruli it is almost not filtered). It provides contrasting images of the functional renal parenchyma, captures the dynamics of urine outflow from the kidneys well, and can be used to determine the effective renal plasma flow (ERPF) by plasma clearance. For this purpose, use is sometimes also ortojodhippuran labeled with 123I (formerly 131I), which is excreted from about 80% by tubular secretion and about 20% by glomerular filtration.
Execution : 
Under the scintillation camera, set in the rear projection on the kidney area, approx. 200MBq of radio indicator is applied and dynamic acquisition is started immediately in short time intervals: perfusion phase approx. 1s/frame - 100 frames, followed by functional phase approx. 10-30s/frame, total acquisition time 30 minutes. The images are stored sequentially in the computer's memory. In the case of visible retention in the hollow system, a diuretic is applied in about 15 minutes - a substance that increases diuresis (intensity of urinary excretion) by affecting transport in various parts of the nephron; furosemide is the most commonly used.
Processing : 
By observing the sequences of images, we can visually observe the entry of the radioindicator into the kidney, its uptake in the parenchyma, transport into the calyx-pelvic system and outflow trough the ureter into the bladder. We can recognize possible abnormalities. For quantitative analysis, on suitable summation images we mark the areas of interest (ROI): the area of the bloodstream (area around the heart), left and right kidneys and their parts (parenchyma, hollow system), tissue background
(on which correction is performed - subtraction). From these areas, the computer then creates curves ("histograms") of the time course of activity in these places.
    The curve from the area of interest ROI of the bloodstream represents the time changes (especially the decrease) of the concentration of the radioindicator in the blood (plasma). The rate of decrease in plasma radioactivity concentration depends on the elimination ability - clearance - of the kidneys, with impaired renal function, the rate of clearance is reduced
(calculation of clearance is described in the passage "Processing the blood-pool curve - clearance" in the book "OSTNUCLINE").
    Nephrographic curves are created from the ROI of the kidneys, from which the time and speed parameters of reaching the maximum, speed or half-time of the decline, are determined. A diuretic test is important in case of delayed excretion or retention of the radiolabel in the kidney: if the decrease in the nephrographic curve did not occur even after the application of the diuretic, this indicates obstructive hydronephrosis.
    By differentiating the activity of the parenchyma and the pelvis, it is possible to decide whether the pathology of the nephrographic curve is caused by a functional disorder of the indicator passage in the parenchyma, or changes in the outflow - dilatation of the hollow system or obstruction of the urinary tract in the kidney. This is exactly done by the so-called deconvolution analysis of nephrographic curves to create transit functions.
   The mathematical description of deconvolution analysis, creation of transit functions, their analysis and interpretation, is in the passage "Deconvolution, transit functions", Fig.3.4.3 in §3.4 "Dynamic scintigraphy of the kidneys" of the book "OSTNUCLINE - Mathematical analysis and computer evaluation of functional scintigraphy".

Mathematical analysis and complex evaluation of dynamic functional scintigraphy of kidneys - MAG3
Evaluation:
After intravenous administration of the radioindicator, the kidneys of the usual shape, size and placement are displayed, without focal changes. The nephrographic curve of the left kidney has a normal course, on the curve of the right kidney we observe a slowdown of drainage and retention, disappearing after diuretics.

Conclusion:
Visual evaluation of sequential images and quantitative analysis of nephrographic curves indicate good function of both kidneys, rapid transit through the parenchyma and free drainage of the hollow system. In the right kidney, a slight slowing of dilatation- type drainage .
                                                                                                                                               Signature: MUDr. Jozef Kubinyi, Ph.D.

Here are examples of evaluation almost normal á and distinctly pathological â dynamic renal scintigraphy.

Mathematical analysis and complex evaluation of dynamic functional scintigraphy of kidneys - MAG3
Evaluation:
After intravenous administration of the radioindicator, a well-accumulating left kidney of the usual shape and size was displayed, without focal changes. The right kidney appears delayed as markedly hypofunctional and inhomogeneous - only the narrow margin of the functional parenchyma around the markedly dilated excavated hollow system is preserved, with significant retention. The nephrographic curve of the left kidney has a physiological course. The nephrogram of the right kidney has a markedly flat shape with a low functional segment, the curve has a permanently ascending course, unresponsive to the application of a diuretic in the 17th minute.


Conclusion:
Visual evaluation of sequential images and quantitative analysis of nephrographic curves indicate good left kidney function, but severely hypofunctional right kidney with marked renal parenchymal atrophy. Left kidney drainage is a physiological, on the right is an obstructive drainage disorder, without response to the administered diuretic. Global kidney function is almost normal due to age.
                                                                                                                                              Signature: MUDr. Jozef Kubinyi, Ph.D.

The mathematical procedure for the analysis of dynamic scintigraphy of the kidneys is described in detail in §3.4 "Dynamic scintigraphy of the kidneys" of the book "OSTNUCLINE - Mathematical analysis and computer evaluation of functional scintigraphy".
Renovascular hypertension - captopril test 
Elevated blood pressure is a disorder that can seriously endanger health, especially vascular complications. It is usually a primary hypertensive disease, but high blood pressure can also be caused secondarily, by a disease of some other organs. This is often secondary to nephrogenic hypertension in kidney disease such as pyelonephritis or glomerulonephritis. Here, nuclear nephrology can also be used in differential diagnosis. A specific case is renovascular hypertension - increased blood pressure caused by insufficient perfusion of the kidneys (their ischemia) due to stenosis of the renal artery. This retains water and sodium in the body, as even otherwise healthy kidneys cannot sufficiently fulfill their function. The RAAS renin-angiotensin-regulatory system is activated: renin, produced to an increased extent in the ischemic kidney, is converted to angiotensin II by the action of a conversion enzyme, which maintains the blood flow of the glomerulus at the required value by increasing the pressure in the glomerulus. However, this physiological compensation by systemic action on the arterioles and an increase in aldosterone levels leads to an undesirable increase in blood pressure. Inhibition of angiotensin converting enzyme (ACE) by a suitable drug can block this regulatory mechanism. Such a short-acting ACE inhibitor is captopril, which can be used here for diagnostic purposes.
  We therefore apply captopril before starting dynamic scintigraphy which, by inhibiting ACE, reduces tonus in vas efferens and reduces glomerular filtration. The secretion of DTPA by glomerular and MAG3 by tubular cells is slowed down, so that the originally normal nephrographic curves become pathological - a decrease in glomerular filtration and thus a slowing down of the transport of the radiopharmaceutical by the renal parenchyma. By comparing the kidney curves from native dynamic scintigraphy without captopril with scintigraphy after captopril, we can reveal the renovascular origin of hypertension.
Dynamic scintigraphy of the transplanted kidney 
The principle and methodological procedure are basically analogous to the above-mentioned dynamic scintigraphy of the kidney. In addition to the assessment of the clearance function of the transplanted kidney, it is important to assess in particular its perfusion, acute tubular necrosis *) and the risk of rejection, drainage of the graft and ureter, detection of possible complications and anomalies in transplantation (such as urinoma).

*) Somewhat misleading name "acute tubular necrosis - ATN" means a delayed onset of perfusion and renal graft function after transplantation, depending on the duration of cold ischemia kidneys in the time interval between removing the kidney from donor and its transplant to the recipient. Severe ATN may result in rejection.
    Compared to the above-mentioned dynamic scintigraphy of the kidneys, the projection differs - in the AP projection, the camera captures the area including the iliac arteries, the transplanted kidney itself and the bladder. Areas of interest are drawn: blood pool, arteria illiaca, transpl. kidney, bladder, background. In addition to the parameters common in dynamic renal scintigraphy, the Hilson perfusion index and bladder outflow rate are determined. Analysis of dynamic scintigraphy of the transplanted kidney is described in §3.5 "Dynamic scintigraphy of the transplanted kidney" OSTNUCLINE book.

Mathematical analysis and complex evaluation of dynamic functional scintigraphy of a transplanted kidney
Evaluation:
After intravenous administration of a radioindicator, the abdominal aorta and iliac artery are imaged in the usual way, followed by a well-perfused transplanted kidney. In the further course, the radiolabel is well concentrated in the transplanted kidney, then excreted into the bladder quickly enough.

Conclusion:
Visual evaluation of sequential images and quantitative analysis of the curves of the passage of the radio indicator indicate good perfusion and function of the graft , rapid transit through the parenchyma and free drainage of the hollow system. There are no signs of incipient rejection.

Static scintigraphy of the kidneys
Purpose: Using this simpler method it is possible to obtain functional-morphological information about the distribution of the functional parenchyma in the kidney, and derivatively about the shape, size and location of the kidneys, sometimes structural changes. Separate renal function can be determined - % share of left and right kidney in total function (renal functional symmetry test).
Radiopharmaceuticals: Labeled substances are used as radiopharmaceuticals that are taken up by the renal parenchyma, but do not pass into the urine. The most widely used is
99mTc-DMSA (dimercaptosuccinic acid), which is taken up in proximal tubule cells.
Execution: After iv application approx. 100MBq
99mTc-DMSA is performed in about 2 hours by its own static scintigraphy, mostly in 4 projections, the most important of which is the projection of PA and AP.
Evaluation is mostly visual, the size, shape and placement of the kidneys, the distribution of functional tissue are evaluated. Computer processing is performed to determine the separated function
(corrected for the absorption of radiation g from various deep-seated kidneys), is described in §3.7 "Static scintigraphy of the kidneys" of the book OSTNUCLINE.
Radionuclide urowlowmetry and cystography 
Purpose:
Used to examine the dynamics of micturition, determination of bladder volume, bladder residue, evacuation rate, detection of vesicoureteral reflux.
Execution:
A patient whose bladder is filled with a radioactive solution is urinated in front of a camera detector, while dynamic scintigraphy of the ureter and bladder area is scanned, at a frequency of about 1 frame/sec. There are two methods of filling the bladder, direct and indirect cystography. The indirect method consists in the application of a nephrotropic radiopharmaceutical (usually 99mTc-MAG3, approx. 200MBq), after which a normal dynamic scintigraphy of the kidneys is performed, during which the bladder is filled with renal function. In direct cystography, the bladder is filled with a radioindicator (approx. 50MBq) through the catheter.
Evaluation: In addition to the visual assessment of a series of images of the micturition, we mark the areas of interest of the bladder, ureters and tissue background, from which we create curves of the time course of radioactivity, especially the urodynamic curve. By their computer analysis we can quantify the course of micturition - determine the duration and speed of micturition, bladder residue. We can visually assess and quantify the regurgitation of urine from the bladder to the ureters or possibly to the renal pelvis - vesicoureteral or vesicorenal reflux . ....
Computer analysis of dynamic uroflowmetry is described in §3.6 "
Radionuclide uroflowmetry" of the book OSTNUCLINE.

Mathematical analysis and complex evaluation of dynamic uroflowmetry
Evaluation:
After the start of dynamic scintigraphy, micturition soon begins, during which there is a sufficiently rapid emptying of the bladder with a low residue. Vesico-ureteral reflux, more pronounced on the right, is well visible in scintigraphic images of the emptied bladder and urinary tract.


Conclusion:
Radionuclide uroflowmetry shows normal micturition flow and low bladder residue, but shows vesico-ureteral reflux .

4.9.3 Diagnostics of the gastrointestinal tract - liver and bile ducts, pancreas, spleen, esophagus and stomach
Liver scintigraphy 
The liver is an important organ in which significant metabolic, detoxification and elimination processes take place; they are incorporated into the digestive tract and also into the reticuloendothelial system (RES). They are one of the largest organs, weighing about 1.5 kg in humans, are located in the right diaphragmatic arch of the abdominal cavity. The liver parenchyma consists mainly of polygonal liver cells - hepatocytes (60%) and reticuloendothelial Kupfer cells (15%). Then there are the hepatic star-shapet It cells, Pit cells and walls of a large number of blood vessels and intrahepatic bile ducts.
    Hepatocytes take up various substances from the plasma, transform them and then excrete them into the bile, which leaves the intrahepatic pathways (via the gallbladder) through the ductus choledochus to the intestinal tract. Liver cells are significantly involved in a number of metabolic and synthetic processes :   
- Carbohydrate metabolism - liver cells take up glucose from portal blood and convert it to lipids or glycogen, conversion of lactate and alanine to glucose.
- Lipid metabolism - fatty acid synthesis and oxidation, glycerol formation, phospholipids, cholesterol, lipoproteins ...
- Amino acid metabolism -
-
Synthesis of plasma proteins -
- Detoxification function - it is mainly the detoxification of ammonia (formed during the decomposition of amino acids). Ammonia is converted to urea and glutamine, which are then excreted in the kidneys. Furthermore, some foreign molecules (especially hydrophobic, which cannot be excreted by the kidneys) are oxidized by cytochrome and excreted in bile or plasma (from where they are then removed in the kidney).
Red blood cells have a limited lifespan (about 120 days), after which they are taken up in the spleen and liver and sequestered. From the hemoglobin the iron is separated, which is used to synthesize new hemoglobin, while the remaining component (heme), bound to the hemopexin protein, is phagocytic by Kupfer cells and converted to bilirubin. It binds to the blood protein albumin and is taken up by hepatocytes in the liver, where it binds to glucuronic acid. This produces conjugated bilirubin, which the liver cells secrete into the bile, which then drains into the small intestine.
The liver also has a number of other functions - the production of hormones and their degradation or inactivation, cholesterol degradation, part of hematopoiesis, storage functions of lipids and glycogen.
  Hepatocytes secrete water, salt ions, acids, cholesterol, phospholipids and bilirubin - liver bile - into the bile capillaries, which gradually coalesce into the bile ducts. Bile collects in the gallbladder - a sac-shaped "reservoir" on the bile, from where it is released in a controlled manner and drained through the bile ducts - ductus cysticus - ductus choledochus - into the duodenum and from there into the intestinal tract (small intestine), where it is involved in fat digestion.
    Kupfer cells are fixed macrophages that phagocytose bacteria, foreign proteins, persistent erythrocytes and some other cells. Pit cells are large granular lymphocytes with significant cytotoxic activity, which by their phagocytic ability cooperate with Kupfer cells. It's cells (lipocytes) contain a large amount of lipids.
    The blood circulation of the liver is about 1.5 liters/minute and has two components :
1.
The hepatic artery supplies blood rich in oxygen (20% of the blood circulation of the liver), nourishes the liver parenchyma.
2.
Functional circulation from the portal vein - vena portae (approximately 80% of the hepatic circulation) brings blood containing absorbed nutrients from the digestive tract, as well as various products of cell metabolism. The vena portae branches into veins flowing through the portobiliary space between the hepatocytes, the blood then drains through the hepatic veins into the inferior vena cava.
Pathology of the liver and bile ducts 
The liver has a large functional reserve and the ability to regenerate. However, the liver can be damaged due to excessive exposure to toxic substances, hepatotoxins (such as alcohol), inflammatory and infectious diseases (hepatitis A, B, C). These damage can result in nodular remodeling, fibrosis, and gradual disappearance of the liver parenchyma - liver cirrhosis associated with liver failure and vascular complications - by portal hypertension and portosystemic shunts (see "spleen" below).
    Pathologies of the bile ducts, especially cholelithiasis, are relatively common - stones in the gallbladder, which are caused by increased concentration and decreased bile solubility. Gallstones can cause inflammation and clog the bile ducts, preventing the outflow of bile from the liver - obstructive ikterus occurs, manifested as jaundice caused by the accumulation of bilirubin in the plasma.
    Liver cancer - the primary liver tumor is hepatocellular carcinoma (hepatoma). Much more common are secondary liver involvement with metastases from other tumors (most often breast or colorectal ca). A benign tumor of the liver is hemangioma.
    Radionuclide diagnostics of the liver primarily uses the functions of hepatocytes and Kupfer cells. After administration of the radiopharmaceutical, which is taken up from the bloodstream by hepatopcytes, we can investigate the function of liver and biliary tract by dynamic liver scintigraphy - cholescintigraphy. By applying a radiopharmaceutical that is taken up (phagocytosed) in Kupfer cells, we can disply the distribution of the parenchyma by static scintigraphy of the liver and thus (indirectly) obtain information about the morphology of the liver.

Dynamic liver scintigraphy - cholescintigraphy
Purpose:

It is used for comprehensive assessment and quantitative analysis of hepatic function (and its parts), clearance and drainage - dynamics of bile formation and outflow through intrahepatic pathways into the gallbladder, gallbladder evacuation, passage to the duodenum and intestinal tract. It can also detect duodeno-gastric reflux. In addition, it may provide some information on the morphology of the liver, which, however, is derived from the display of the radioindicater distribution in the functional tissue of the liver (parenchyma) and from the outflow or accumulation of the radiolabel in the bile ducts.
Radiopharmaceuticals :
Iminodiacetic acid (IDA) derivatives labeled
99mTc - HIDA, EHIDA.
Execution:
After i.v. application of approx. 100-200 MBq of hepatotropic radiopharmaceutical in the supine position in the front projection, dynamic scintigraphy is immediately started at a frame rate of approx. 20-30 sec./frame. To stimulate gallbladder emptying, a cholekinetic stimulus - cholecystokinin or a fatty diet (chocolate) is given at about 30th minute during the examination. Dynamic scintigraphy is scanned for 60 minutes. If the radio indicator does not appear in the intestines after this time, further still images are recorded in 2 and 4 hours.
Evaluation :
By visual evaluation of images of different phases of radioindicator distribution in the liver and passage through the bile system, we assess the distribution of hepatocytes in the liver parenchyma, bile duct morphology, gallbladder deposition and size, its emptying and bile drainage into the intestinal tract, event. duodenal-gastric reflux.
    On the relevant dynamic images, we mark the regions of interest: bloodstream, whole liver, liver parenchyma, ductus choledochus, intestinal tract, gallbladder (if seen) and stomach (if duodenal-gastric reflux is suspected). From the curves of the time course of radioindicator concentration in these areas of interest, we evaluate and quantify hepetocellular liver function - clearance and rate of extraction of the radiopharmaceutical by hepatocytes from the bloodstream, biliary outflow dynamics, including determination of gallbladder ejection fraction, degree of duodeno-gastric reflux.
    By deconvolution of the liver curves with the blood-pool curve, we can construct transit functions and determine the transit times of the passage of the radio-indicator through the entire liver, liver parenchyma or its selected parts. By exponential analysis of transit curves, we can further determine the hepatic extraction fraction (HEF) of the liver parenchyma, its individual lobes or selected parts (ROIs). HEF indicates how much [%] of the radio-indicator is absorbed during one passage through the vascular bed of the liver parenchyma - thus it quantifies the liver parenchyma's own functional absorption ability, which enables it to be distinguished from biliary tract pathology.
    The method of determining HEF is described in the passage "Hepatic extraction fraction" §3.10 "Dynamic scintigraphy of the liver" of the book "OSTNUCLINE".
    Determining the functional capacity of the liver and its individual segments and selected areas is important not only in clinical hepatology, gastroenterology, internal medicine, but also in patients undergoing more extensive resection procedures with the removal of part of the functional liver parenchyma. Here there is a risk of insufficient function of the remaining functional residue and the development of liver failure, with very adverse
(even fatal) health consequences. Mapping the functional fitness of the liver parenchyma can contribute to the optimization of the planned surgical strategy.

Mathematical analysis and complex evaluation of dynamic functional scintigraphy of the liver and bile ducts - cholescintigraphy
Evaluation:
After intravenous administration of the radioindicator, the liver of the usual shape and size is imaged in a timely manner. The liver parenchyma does not show focal changes. Bile ducts can be differentiated from 10th minutes, from 13th minutes the gallbladder begins to fill. In 30.min. a cholekinetic stimulus was administered. In the next course, we observe a rapid passage of the radioindicator through the biliary system, with a smooth outflow through the ductus choledochus into the intest. tract.


Conclusion:
Visual evaluation of sequential images and quantitative analysis of liver curves indicate good hepatocellular function, rapid transit through the liver parenchyma and free drainage of the biliary system, without signs of biliary obstruction. The gallbladder has a good filling and evacuation function.

Mathematical analysis and computer evaluation of dynamic cholescintigraphy is described in detail in §3.10 "Dynamic liver scintigraphy" of the book "OSTNUCLINE".

Diagnosis of bile acid malabsorption
Bile acids are synthesized in the liver from cholesterol. The primary bile acids, cholic and chenodyoxycholic, are conjugated with glycine and taurine and are expressed into the bile ducts. After the intake of fat in food, they are secreted into the duodenum under the influence of cholecystokinin. Bile acids have a decisive function in the digestion and absorption of fats taken in food - fatty acids and momoglycerides released during lipolysis in the intestine. Bile acids with this process are gradually absorbed in the small intestine and do not reach the large intestine.
    With malabsorption, bile acids are not sufficiently absorbed in the ileum, they penetrate into the large intestine and are the cause of chronic diarrhea. They cause increased secretion of electrolytes and water, increased permeability of the colon mucosa, increased colon motility and increased mucus production. This disorder of reduction in the absorption function of the terminal ileum occurs, for example, in Crohn's disease, inflammatory, toxic or radiation damage; it often occurs after radiotherapy of malignancies located in the lower parts abdomen and in pelvis.
Radiopharmaceuticals :
The SeHCAT (Selenium HomoCholic Acid Taurine Test) method for diagnosing bile acid malabsorption uses
75Se- tauroselcholic acid as a radioindicator. It is a synthetic bile acid conjugated with taurine (23-seleno-25-homo-taurocholic acid), labeled with radioactive selenium 75Se. Tauroselchic acid is similar to bile acid, it has the same physiological behavior as conjugates of bile acids.
The physical properties of selenium-75 are given in §1.4 "Radionuclides", passage "
Se-75". Measurement of radiopharmaceutical retention kinetics was previously performed experimentally using whole-body radiometers or collimated scintillation probes. In nuclear medicine workplaces, the use of a scintillation gamma camera without a collimator proves to be the best - due to the low applied activity, we need maximum detection sensitivity, but we do not need a scintigraphic display of the radioindicator distribution, we just need to measure the total number of impulses from gamma radiation in a given time -> relative activity of the radiopharmaceutical.
Execution :
 Approximately 400 kBq of 75Se-tauroselcholic acid is administered orally (in a capsule) to the patient. Under the camera without a collimator, placed above the abdomen, the number of pulses registered for an acquisition time of about 10 minutes is measured, the window of the analyzer with a width of 20% is set to the photopeak of 289 keV 75Se. The first measurement is performed 3 hours after application (day 0), the second measurement is performed 1 week later, in the identical position of the patient (on the back, same area of the abdomen). In both measurements, the radiation background is subtracted, which tends to be relatively high; measurements need to be made under conditions where there are no radioactive sources in the vicinity (e.g. other patients with applied radioactivity). A correction is made for the decay of Se-75 - in 7 days it is a coefficient of 1.04. From the values measured in this way, we determine the retention of the radiopharmaceutical in % (the value on the 7th day is expressed as a percentage of the value on the 0th day). A normal rate of loss is >20% per week. Retention of 10-15% is considered mild malabsorption, 5-10% moderate, <5% is classified as severe malabsorption.
    This method is relatively marginal in nuclear medicine, it is performed only rarely only in some workplaces. A therapeutic test of the clinical response to sequestrants after administration of cholestyramine and biochemical tests are usually sufficient for gastroenterologists. Nevertheless, in some more complex cases, this special examination can be useful...

Static scintigraphy of the liver
Purpose: To obtain indirect information about the morphology of the liver - findings of diffuse involvement and detection of focal liver lesions.
Radiopharmaceuticals:
Sn-colloid (or sulfur colloid) labeled with 99mTc, which is rapidly taken up from the bloodstream by Kupfer cells after application.
Execution: After i.v. application of approx. 150 MBq 99mTc-Sn colloid, in 15 min. performs scintigraphy in the front, back and possibly right lateral projection. For better imaging of lesions deposited deeper in the parenchyma, it is advisable to perform SPECT imaging.
Evaluation:

In addition to the placement, shape and size of the liver, we visually assess the distribution of the radiopharmaceutical in the parenchyma on planar or tomographic images. Liver lesions are usually accompanied by decreased Kupfer cell density, which results in decreased radiopharmaceutical accumulation. The finding is non-specific : local reductions (cold deposits) may be caused by cysts, abscesses or tumors (metastases), diffuse involvement (hepatomegaly, uneven distribution in the parenchyma, increased accumulation in extrahepatic RES - spleen and bone marrow) may be caused by hepatitis, cirrhosis, metabolic disorders, malignancies.
Scintigraphy of hepatic hemangiomas 
Hemangioma
is a benign mesenchymal tumor of blood vessels. A cavernous type of hemangioma often occurs in the liver. It is a highly vascularized structure that has a higher proportion of blood - and thus a higher concentration of erythrocytes - than the surrounding tissue. After application of radionuclide-labeled erythrocytes, hemangiomas appear as "hot" deposits of increased radioactivity deposition than in the surrounding tissue.
Purpose: The examination is used to detect cavernous hemangiomas in the liver and to distinguish them from other structures (such as primary or metastatic tumors).
Radiopharmaceuticals: Autologous erythrocytes labeled with 99mTc - labeled either in vitro, but more often in vivo using Sn-pyrophosphate (applied 20 min. before application of 99mTcO4).
Execution: Simultaneously with the i.v. application of the radiopharmaceutical, dynamic scintigraphy of the perfusion phase *) is started in a projection in which the best imaging of suspicious deposits is assumed. Total dynam. shooting about 2 min., frame rate 2-3 s./frame. After 40-60 minutes, static scintigraphy of the liver area is performed, preferably in SPECT tomography.
*) Note: Previously performed dynamic scintigraphy of the perfusion phase in hemangiomas it is based on the fact that in the hemangioma there is an increased blood pool at a relatively slower blood flow compared to the vessels of the surrounding tissue. However, monitoring of the perfusion phase has been shown to have little clinical benefit and is therefore generally not performed.
Evaluation:
On static planar scintigrams and on reconstructed tomographic SPECT sections, we look for deposits of increased deposition of labeled erythrocytes, which indicate the presence of hemangiomas. In SPECT scintigraphy, the limit of detection of hemangiomas is about 1 cm.
    Distinguishing hemangiomas from other units suspected of malignancy is important for primary tumor diagnosis. Biopsies should not be performed on hemangiomas as highly perfused structures, because there is a risk of bleeding .
Pancreas scintigraphy 
The pancreas is a small but metabolically and endocrine important organ located in the abdominal cavity in the duodenum, just below the liver. The exocrine component, which opens into the duodenum, produces digestive enzymes - pancreatic lipase for the breakdown of fats, alpha amylase for the breakdown of starch, proteases for the breakdown of proteins. Under normal circumstances, digestive enzymes are inactive form after their formation inside the pancreas (otherwise they would damage - "digest" - the pancreatic tissue, pancreatitis would occur), only when they reach the duodenum are they activated and can begin to perform their digestive function. Endocrine part (whose cells are arranged in the islets of Langerhans) produces pancreatic hormones - insulin (regulates blood sugar levels), glucagon, somatostatin, pancreatic polypeptide.
Pancreatic pathology 
The most common disease associated with pancreatic is diabetes - diabetes mellitus caused by tissue damage islets of Langerhans, where insulin is formed. Inflammation of the pancreas, pancreatitis, is caused by the retention of digestive enzymes in the pancreas, which remain inside, are prematurely activated and "self-digest" damage the pancreatic tissue, causing swelling and an inflammatory reaction. 
Acute pancreatitis is caused by sudden obstruction of the pancreatic duct to the duodenum, usually by bile stones. Chronic pancreatitis, caused by a slow outflow of pancreatic enzymes into the duodenum, has a milder and longer-lasting course. In more severe cases of necrotizing hemorhagic pancreatitis, proteolytic pancreatic enzymes can enter the bloodstream and cause toxic effects in various tissues and organs. A very serious disease is (adeno)carcinoma of the pancreas, which often metastasizes the whole body through the lymphatic system.
Dynamic scintigraphy of the pancreas 
The pancreas is difficult to access for functional diagnosis. For scintigraphic examination of the pancreas the radiopharmaceutical 75Se-selenomethionine H2 C-S- (CH2 )2 CH (NH2 ) COOH labeled with the radionuclide selenium 75Se was developed (physical properties in §1.4 "Radionuclides", passage "Se-75 "). The intake of amino acids in the pancreas is a reflection of the rate of synthesis of digestive enzymes. The similarity between selenium and sulfur is so close that the substitution of selenium instead of sulfur in the methionine molecule leads to an analogue that has all the metabolic properties of an amino acid, including incorporation into proteins, and is therefore efficiently taken up by the pancreas in digestive enzyme production.
    The first attempts at radioisotope examination of the pancreas in the 1960s with collimated probe detection ("blind" measurement - a priori to no avail...) and then static scintigraphy with a motion gammagraph and camera without computer acquisition, were able to assess only gross pancreatic abnormalities. Valid results were obtained only by dynamic scintigraphy of the pancreas :
    Dynamic scintigraphic examination of the pancreas can be useful for early detection of pancreatic exocrine dysfunction, retention or obstruction of the drainage pathways - in diabetic patients, pancreatitis, cancer.
    Gamma camera equipped with a ME or HE collimator, the analyzer window set to a 264keV peak, is placed the slightly obliquely above the liver and pancreas area. After i.v. application approx. 100 kBq/kg 75Se-selenomethionine, dynamic sequential images the liver and pancreas are taken on a gamma camera at intervals of 5-10 minutes. for 60-120 minutes. Selenomethionine is also taken up non-specifically in the liver (where proteosynthesis also takes place), so that the image of the pancreas is often displayed in interfering background against the radioactivity of the liver. To eliminate this disturbing background, 99mTc-colloid is sometimes applied at the end of the examination (with the patient's position unchanged), which is specifically taken up by the liver. The resulting scintigraphic images of the liver are then subtracted from the 75Se-selenomethionine images. By this gradual subtraction of images, the disturbing image of the liver is suitably suppressed and better separation and visibility of the pancreas is achieved.
    The regions of interest (ROI) of suitable parts of the pancreas are then marked on scintigraphic images, from which dynamic curves of the time dependence of selenomethionine accumulation are generated. The curves are mostly evaluated visually (but at our workplace we also developed a program for their quantitative processing). In physiological cases, the curve after the initial rapid increase reaches a peak after 20-30 min. from the application, followed by a slower decline. Reduced function of the pancreas is manifested on the curve by a flat shape, with a later onset and slowing of the rate of growth. In this case, the pancreas is displayed less clearly on scintigraphic images. Retention of digestive enzymes within the pancreas is manifested by a later and slower decline; with more severe pathology, the peak and decline do not appear at all, the curve still has a slowly increasing trend.

Dynamic scintigraphy of the pancreas.
Left:
The area of interest (ROI) of the appropriate part of the pancreas is marked on the summary scintigraphic image. Middle: Typical normal and pathological curves of the time course of selenomuthionine uptake in the pancreas. Right: Different responses of the curves to cholecystokinin application.
(This scintigraphy was taken on a Clincom instrument)

In such pathological cases, in about 60.-90. minute i.v. applies pancreozymin, more commonly called cholecystokinin (also stimulates gallbladder contraction), 1 u/kg, which stimulates the secretion of pancreatic enzymes (and possibly also secretin, which, among other things, potentiates the effect of cholecystokine and also has trophic effects on the pancreas). It is analogous to the above cholekinetic stimulus in dynamic cholescintigraphy, or the application of a diuretic (furosemide) in dynamic renal scintigraphy. Depending on the response of the curve to this pancreatic stimulator, it is possible to distinguish the decrease in function, parenchymal pancreatic damage or obstruction ...
  Pancreatic scintigraphy was relatively infrequent, diagnostic yield relatively low, with a significant percentage of indeterminate findings. It was performed mainly in the 70s-80s. Pancreatic cancer is now visualized by CT imaging and static PET/ CTscintigraphy.
  Static scintigraphy of the pancreas with 99mTc- interleukin-2 is also tested for imaging of chronic inflammatory changes in type 1 autoimmune diabetes, to identify patients with pancreatic inflammation. It makes it possible to detect an increased incidence of activated T cells - the degree of lymphocytic infiltration even in smaller inflammatory processes of insulitis, in the early period for the treatment of immunotherapies.

Scintigraphy of the spleen and dynamic splenoportography
The spleen
(Latin lien , Greek splén ) is a somewhat "mysterious" organ located in the abdominal cavity, near the stomach. Phylogenetically, the spleen probably developed as an organ of hematopoiesis. However, it retains this function only in the prenatal period (until about the 6th month of fetal development), then it is taken over by the bone marrow. After birth, the spleen functions only as a "filtering" organ with a large number of macrophages, retaining microorganisms from the blood and obsolete or damaged ("worn out") blood cells - sequestration of erythrocytes. It is also has immune significance, produces antibodies and immunocompetent cells, has phagocytic ability (RES system).
    The weight of the spleen is about 100-200g, gradually decreasing with age. In some diseases, however, the spleen enlarges - spenomegaly. Mild splenomegaly (weight up to 500g) can also occur during infections. Moderate splenoagaly (500-1000g) may accompany acute leukemia, malignant lymphoma, polycythemia and more. Rarely, severe splenomegaly (weight > 1000 g) occurs, eg in chronic myeloid leukemia, ...
    Anatomically, the spleen belongs to the reticuloendothelial (RES) and hematopoietic system. The blood supply to the spleen takes place through the portal circulation (vena portae*), thus being significantly connected to the liver. Therefore, we have placed scintigraphic methods related to the spleen in the context of liver scintigraphy.
*) The portal vein drains blood from the organs of the abdominal cavity - from the intestines, lower esophagus, stomach, spleen, pancreas - to the liver.
Pathology of the spleen and portal pathways
 
One of the pathologies of the spleen is the above-mentioned splenomegaly. Portal hypertension is increased blood pressure in the basin of the portal vein (vena portae). The portal vein block is most often intrahepatic due to liver cirrhosis. Instead of the portal vein, the blood then flows through the created "connectors" - portosystemic short circuits - into the systemic circulation (into the basin of the inferior vena cava). There is a development collateral flow, overloading of the veins creates varices. Blood from the digestive system bypasses the liver, so it is not detoxified, which can lead to damage to some tissues (e g brain).
Increased destruction and sequestration of erythrocytes
in the spleen is manifested in hemolytic anemia
(see "Half-life of erythrocytes and localization of their destruction"). This often leads to splenomegaly. In this case, splenectomy is recommended to normalize the blood count.
Static scintigraphy of the spleen 
Purpose:
Imaging of the functional tissue of the spleen to determine its shape, size and placement, including possibly inhomogeneities. It can also be used to visualize and mark the spleen for application before dynamic splenoportography.
Radiopharmaceuticals: 99mTc-labeled autologous heat-damaged erythrocytes are used for selective imaging of the spleen and are replicated to the patient. Radiocolloids, mainly 99mTc-sulfur-colloid, whose larger colloidal particles are taken up in the reticuloedothelium, are used to image the reticuloendothelial system of the spleen (+ liver).
Execution:
After application of about 100-200MBq of the above radio indicator, planar scintigraphic images in the front, back, and left side projections are taken in about 20-30 minutes. For a more detailed distinction of pathological structures, we can add the display of SPECT.
Evaluation: In the pictures we assess the shape, size and placement of the spleen; the size of the spleen can be estimated from the dimensions of the scintigraphic image using empirical methods. By observing the homogeneity of the distribution of the radiopharmaceutical or the presence of focal changes, we can infer abscesses, cysts, splenomegaly, hematomas or tumors in the spleen.
Dynamic splenoportography 
Purpose:
Examination of blood flow through the portal vascular bed and detection of portosystem shunts. Splenoporography is one of the less frequent scintigraphic examinations, now it is almost abandoned ...
Radiopharmaceuticals: 99mTc pertechnetate.
Execution:
The application of the radioindicator is performed intrasplenically with a thin needle in a small volume (up to 10 ml.) and fast enough (bolus) so that the phase of the first flow is well expressed. At the same time, we will launch the acquisition of dynamic scintigraphy in the front projection - 60 images after 2 seconds, which captures the flow of the radio indicator through the portal and system streams.
Evaluation:
Visually evaluate images capturing individual phases of passage: spleen ® v.lienalis ® v.portae ® liver ® systemic circulation. Under normal circumstances, after itrasplenic application, the radiolabel passes rapidly through the v.lienalis and v.portae into the liver, where the flow slows down appropriately in the capillary bed, then flows through the hepatic veins and inferior vena cava into the heart and lungs, and then into the systemic circulation. In the presence of shunts (connectors) of the portal and systemic flow, part of the radioindicator passes out of the liver and reaches the heart prematurely. In addition to these portosystemic shunts, we can also asses the possible obstruction of the v.lienalis or v.portae in a series of images.
    To quantify the dynamics of the flow of the radioindicator through the portal and systemic streams, we mark the relevant areas of interest: vena lienalis, liver, heart + lungs, from which we create curves of the time course of the passage of the radioindicator. From these curves we can quantify the flow dynamics. For curves from the v.lienalis, liver, and heart regions, the time of arrival of the radiolabel, the time of maximum, the steepness (gradient) of increase (the ascending section intersects the linear function) and the half-time escape of the radiolabel (the descending section interpolates the exponential function) are determined.
The procedure for computer evaluation of dynamic splenoportography is described in §3.16 "
Dynamic splenoportography" of the book "OSTNUCLINE".

Evaluation of dynamic splenoportography
Evaluation:
After intrasplenic application of 99m-Tc, we observe on scintigrams a fast flow of the radioindicator through the v.lienalis and v.portae to the liver, without obvious portosystemic short circuits. After the usual slowing down in the capillary bed of the liver, the radiolabel flows out through the inferior vena cava into the heart and lungs.

Conclusion:
Visual evaluation of sequential scintigrams as well as quantitative analysis of the curves indicate normal flow conditions in the portal stream, without obvious portosystemic shunts.

Here are examples of the evaluation of normal á and significantly pathological â radionuclide splenoportography.

Evaluation of dynamic splenoportography
Evaluation:
After intrasplenic application of 99m-Tc, we observe on scintigrams a fast flow of the radioindicator through the v.lienalis to caudal shunts. The liver is practically invisible. Long-term retention of the radioindicator in the spleen. Through caudal protocaval shunts, the radioindicator flows into the heart and lungs.

Conclusion:
Visual evaluation of sequential scintigrams and quantitative analysis of the curves shows significant caudal portosystemic shunts, with no apparent flow through the portal vein to the liver.

 Scintigraphy of the esophagus and stomach
Esophagus used to swallow food (solid and liquid phase) from the mouth to the stomach, where there is a first stage digestion of food. Under a physiological state, the swallowed bite is actively transported by the peristalsis of the esophagus to the stomach, where it arrives in about 7 seconds. The motility disturbances of the esophagus may occur due to stenosis of the esophagus, innervation disturbances, .. Food passage is then decelerated and is irregular. A disorder of the lower esophageal sphincter causes part of the food to return from the stomach back to the esophagus - gastroesophageal reflux occurs.
Dynamic esophageal scintigraphy - swallowing act
Purpose:
Assessment of esophageal motility, its patency, course of swallowing and detection of the presence and severity of gastro-oesophageal reflux.
Radiopharmaceuticals: 99mTc Sn-colloid or 99mTc-DTPA.
Procedure:
Orally administer about 50 MBq of radioindicator mixed with about 10 ml. water (or fruit juice) and immediately (better with a little advance) we start a fast dynamic scintigraphy sitting in the front projection - 120 images after 0.5 sec. (captures the passage through the esophagus) and then about 60 images after 30 sec. (captures gastric evacuation or late reflux). During the examination we can possibly. perform a compression on the epigastrium or Valsava maneuver to provoke gastroesophageal reflux.
Evaluation:
First, we visually observe images of the passage of the radioindicator through the esophagus, its distribution in the stomach and then its gradual evacuation to the intestinal tract. Under normal circumstances, a swallowed bite is rapidly transported to the stomach with the help of esophageal peristalsis, so that the passage of the radioindicator through the esophagus must be sufficiently rapid and smooth, without temporary or permanent retention. In various pathological conditions such as achalasia, disorders of patency (narrowing of the lumen of the esophagus - tumor, external oppression of the esophagus, etc.), or disorders of esophageal innervation, disorders after operations on the esophagus, the passage through the esophagus slows down. In scintigraphic images, we then see slowed down or uneven passage through the esophagus, which may be accompanied by retention of the radioindicator in some parts of the esophagus. However, only more pronounced abnormalities are seen in the scintigraphic images; more detailed and sensitive analysis and quantification of the esophageal passage is performed on the curves from individual parts of the esophagus and on special mathematical constructions - transport function and condensed image.
    A common pathology is gastroesophageal reflux, when due to insufficiency of the lower esophageal sphincter,  part of the gastric contents return to the esophagus, ie abnormally oriented movement against physiological direction of food passage. In the relevant scintigraphic images, the resulting regurgitation manifests itself as the presence of a radioactive deposit, especially in the area of the lower third of the esophagus (reflux can extend also to the higher levels of the esophagus - more detailed and sensitive analysis of the presence and location of reflux is performed on curves from individual parts of the esophagus). Reflux can occur either passively (spontaneously, under native conditions), or it can be caused by increased pressure in the stomach (manifested by appropriate compression of the stomach area) - then it is active reflux.
    Analysis and computer evaluation of dynamic scintigraphy of esophageal swallowing function and gastric evacuation is described in detail in §3.20 "
Dynamic scintigraphy of the esophagus and stomach" of the book "OSTNUCLINE".

Mathematical analysis and complex evaluation of dynamic esophageal scintigraphy - swallowing act
Evaluation:
After oral administration of the radioindicator, we observe in scintigraphic images first a rapid passage of the upper and middle part of the esophagus, then a somewhat slowed passage of the distal part of the esophagus. Once the stomach is reached, most of the radiotracer returns to the middle stage of the esophagus, where it retains for about 20 seconds and only then progresses to the stomach. This abnormal movement of the swallowed radio indicator is particularly evident in the condensed image and the transport function.

Conclusion:
Visual evaluation of sequential images and quantitative analysis of radioindicator passage curves indicate good patency and motility of the upper esophagus, while severe passage pathology with marked gastro-oesophageal reflux was observed in the middle and lower esophagus (probably related to incopetence of the cardia).

Dynamic scintigraphy of evacuation of the stomach and small intestine
Purpose:
Monitoring the rate of evacuation of food from the stomach to the intestine, or the rate of transport trough the small intestine.
Radiopharmaceuticals: 99mTc Sn-colloid or 99mTc-DTPA, which we mix into solid or liquid food.
Execution:
To a patient sitting in front of the camera, we give orally a small bite of solid food, marked approx. 50 MBq 99mTc. We will start dynamic scintigraphy, in which we scan the stomach area for about 1.5 hours at a frequency of 1 frame per minute. If we want to investigate the evacuation of liquid food, dynamic scintigraphy of the stomach may follow dynamic scintigraphy of the swallowing act of the esophagus (described above). However, a solid diet is more representative for assessing gastric evacuation. In addition, scintigraphic examination of small bowel transport may follow, in the form of a slow dynamic study or sequential still images; it is taken for several hours as needed. It is advisable to take a control image of the abdominal cavity the next day (after 24 hours).
Evaluation: On the images of the stomach, we mark the ROI, from which we generate a curve of the time course of the activity. This curve begins with a flat arm corresponding to a phase in which the ingested diet does not leave the stomach (or leaves it only very slowly). This is followed by a various fast decrease in activity, capturing the evacuation from the stomach to the intestine. We evaluate half-time of the evacuation of the stomach T1/2 - the time required for the activity in the stomach to decrease by half (the descending part of the curve interpolates the exponential function, from the rate coefficient of which we determine T1/2). Normal values of the half-time of gastric evacuation in the case of solid food are in the range of 60-90 min., in the case of liquid food approx. 30-40 min. If scintigraphy of the small intestine followed, we determine in the sequential images the time since ingestion, for which the activity first appears in the initial wide part of the large intestine (caecum). It is the so-called oro-caecum time, whose normal values are about 2-5 hours. As the half-time of gastric evacuation also affects the rate of small bowel transport, it must also be taken into account.

Scintigraphic localization of bleeding into the GIT
It is performed using labeled erythrocytes. We take about 2-5 ml. blood, in which
99mTc is labeled with erythrocytes in vitro and reinjected back. After i.v. application of approximately 500 MBq of these in vitro labeled autologous erythrocytes is performed the abdominal scintigraphy. First dynamic scintigraphy about 60 images after 1 min. If extravasal activity does not appear on the images by then, we continue by acquisition of static images at approximately hourly intervals. Possibly bleeding is reflected in the scintigrams by an increase in the concentration of radioactivity in the area where the bleeding occurs.

Scintigraphy of Meckel's diverticulum - imaging of ectopic gastric mucosa
Diverticulum in medical terminology generally refers to local bulging of a hollow organ wall. Meckel's diverticulum
(in the narrower sense) is the bulging of the wall of the small intestine formed by the ectopic gastric mucosa. It can be scintigraphically imaged with 99mTc-pertechnetate, which is physiologically taken up by gastric mucosal cells. After i.v. application of this radiopharmaceutical, we perform dynamic imaging of the abdomen, approx. 1 min./frame for 1 hour. Simultaneously with the display of physiological accumulation in the gastric mucosa, a possible district of the ectopic mucosa of the Meckel's diverticulum is also displayed.

4.9.4 Nuclear cardiology
The heart
(Latin cor , Greek cardia ) is a hollow muscular organ that, with its regular contractions, functions as a pump that drives blood circulation throughout the body. This ensures the transfer of respiratory gases, nutrients and metabolic waste products. Cardiology deals with the structure, function and diseases of the heart. The heart of higher organisms, especially mammals and humans, consists of several anatomical and functional parts :
- Cardiac cavities and supply vessels
Deoxygenated blood (passed through the organism) is supplied to the heart through hollow veins - upper and lower, which connect to the venous canal in front of the heart. During the flow through the heart, the blood passes through 4 cavities, which are separated from each other by valves, preventing the backflow of blood. Blood flows from the venous canal into the right atrium. From there, it enters the right ventricle through a tricuspid valve. The right ventricle, with its contractions, expels blood through the "crescent" valve into the lungs - the main arteries of the pulmonary circulation. As it passes through the lungs, the blood is oxygenated. Oxygenated blood flows from the lungs through the pulmonary veins into the left atrium and from there through the bicuspid valve (also called mitral for resemblance to the shape of a bishop's miter) into the left ventricle. With the contractions of the left ventricle, blood is expelled through the aortic valves into the aorta, whereby oxygenated blood enters the main arterial circulation - it passes through individual tissues and organs, releases oxygen, transports nutrients, receives metabolic products and returns to the heart via venous system.
    The "pumping" of blood takes place by alternating the phases of systole and diastole of the heart chamber. In systole, the heart chamber contracts and blood flows from the heart chambers into the arteries. During the relaxation phase - diastole - the muscles of the ventricles weaken and the heart fills with blood with passive pressure. Each systole expels about 70 ml from the heart. blood (so-called stroke volume). The amount of water that the chamber pumps per minute is called volum minute heart - cardiac output.
- Heart valves
act as one-way valves that allow blood to flow in only one direction, while closing in the opposite direction and blocking the backflow. There are 4 valves in the heart: - A double- valve (mitral) valve between the left atrium and the left ventricle; - Tricuspid (3 spikes) valve between the right atrium and the right ventricle; - Aortic valve at the interface of the left ventricle and aorta; - Crescent pulmonary valve in the right ventricle in the lung. In order for the valve to function properly, a sufficiently large opening for blood flow must be created when it is opened, and when it is closed, it must fit snugly to prevent blood flow back. A common disorder is the insufficiency of the valves, when part of the expelled blood returns - the so-called regurgitation, during which the heart must then pump it again. This reduces the pumping efficiency and the heart is overloaded.
- Heart muscles
The driving element of pumping is the heart muscle - the myocardium, which drives the heart's pumping activity with its regular contractions. It is a transversely striated, highly powerful muscle. They are made up of cardiac cells by cardiomyocytes. The strongest heart muscle is in the left ventricle, which must expel blood into the great circulation under considerable pressure.
- The vascular supply of the heart
In order for the heart muscle to work, it needs oxygen and nutrients. The vascular supply of the heart muscle with oxygenated blood is provided by two coronary coronary arteries emanating from the aorta. They branch into a network of vessels that surrounds the myocardium and resembles a wreath in shape.
When some sections of the coronary arteries are narrowed (mainly due to atherosclerotic plaques or embolizations), the vascular supply of the heart muscle is reduced - ischemic heart disease. Ischemic necrosis - myocardial infarction - occurs after 20-40 minutes with complete closure of the artery, in which irreversible death of the heart muscle occurs in the basin of a closed vessel.
- Control of cardiac activity
The contraction of the heart muscle is stimulated by electrical impulses. The control of heart activity is largely autonomous - electrical stimuli for myocardial contraction are generated and conducted in the heart wall, in the cardiac conduction system. The main source of excitement is the sinoarthritic node - a cluster of cells in the wall of the right atrium near the venous canal. This node is affected by the autonomic (vegetative) nervous system from the cardioregulation center in the brainstem, in the elongated spinal cord (hypothalamus). The signal is divided into two Tawar arms in the interventricular septum, right and left, which faces the myocardium and spreads excitement along the walls of the ventricles. These electrical excitation signals can be sensed using an ECG.
Cardiovascular pathology 
   Ischemic heart disease consists of a narrowed lumen of the coronary arteries of the myocardium due to atherosclerosis, which results in impaired perfusion of the heart muscle. Severe reduction in perfusion is manifested by angina pectoris, complete closure leads to myocardial infarction . ......
   Defects of the heart valves consist either of a narrowing (stenosis) or of their insufficiency, especially of the mitral or aortic valve. This leads to backflow - regurgitation, which reduces the efficiency of the heart's pumping function. Valves can be affected as part of birth defects, but also as an acquired disability in infectious endocarditis.
   Heart rhythm disorders, also called arrhythmias, can be caused by a disturbance in the production of electrical arousal, or a disturbance in the propagation of arousal. More severe arrhythmias are corrected using a pacemaker.
   Disorders of myocardial contractility - hypokinesia, akinesia, asynchrony, dyskinesia (or aneurysm) .....
   Intracardiac shunts are openings - defects - in the heart wall (septum) between the ventricles or atria. ......
    In connection with the above-outlined function of cardiac activity and its disorders, cardiological diagnostics performs in three basic directions :
1. Acoustic diagnostics of cardiac echoes of systolic-diastolic function using a stethoscope and diagnostics of electrical activity of the heart using electrocardiography ECG. They are the oldest cardiological methods. They are now being approached by ultrasound sonography.
2. Diagnosis of central hemodynamics - measurement of blood flow through the heart cavities and large vessels, detection of intracardiac shunts and heart valve insufficiency, including assessment of their severity.
3. Diagnosis of myocardial perfusion - ischemic heart disease, ischemic myocardial viability ....

  Nuclear medicine can offer cardiology four diagnostic circuits :
- Methods examining systolic-diastolic function of the heart as "pumps" can demonstrate overall and regional impairment of heart wall motility or synchronization with electrical activity of the heart, determine the overall "performance of the heart pump". It is a equilibrium ECG-gated ventriculography and SPECT of the myocardium.
- Examination of central hemodynamics - blood flow through the heart cavities and large vessels. After the application of the bolus of the radioindicator, it is possible to monitor the dynamics of blood flow through large vessels, filling of atria and ventricles, including the detection of incacardiac shunts, to determine the cardiac output, cardiopulmonary blod volume, flow times and other important hemodynamic parameters.
- Examination of the regional blood flow of the myocardium, at rest and under load, allows to diagnose ischemic heart disease, its location and severity.
- Verification of myocardial viability damaged by ischemia. It is important for planning revascularization procedures (by-pass, angioplasty) - revascularization only makes sense in the case of a viable myocardium (which is perhaps only temporarily hibernated by ischemia), not in the case of an already unviable (necrotic) myocardium.

Equilibrium gated ventriculography
Purpose:
It is a dynamic scintigraphic method that provides comprehensive information about the activity of the heart as a pump of blood circulation. Changes in activity in individual cardiac compartments - chambers and atria during their pulsation in the cardiac cycle are displayed. Because the radioindicator is evenly and stably "mixed" in the bloodstream, changes in activity - and thus in the emitted radiation
g - are directly proportional to changes in the volume of the ventricles and atria during pulsation. We can determine the hemodynamic functional parameters of the heart chambers, display the regional kinetics of the heart walls, determine the regurgitation fraction of the left ventricle (along with radiocardiography). The method is useful for assessing the impact of ischemic heart disease or myocardial infarction, or cardiomyopathy, on ventricular function. It is also used to determine the cardiotoxicity of cytostatics in chemotherapy of cancer.
    In continuous dynamic imaging, the number of pulses accumulated during one cycle would be too low to imaging the shapes and sizes of the ventricles and determine their volume changes. Therefore, the ECG gating technique is used: in addition to scintigraphic pulses, an electrical ECG signal is also captured from the camera, which appropriately controls (triggers, "gates") the acquisition process. Gating pulses are derived from the R-wave of the ECG and synchronize periodic storage of scintigraphic images in defined areas of computer memory. Gradual addition of corresponding images from individual cardiac cycles creates the resulting set of images, which represents the phase dynamic scintigraphy of one "representative" cardiac cycle, created by synchronous summation of several hundred ongoing cycles
(described in detail in §4.4 "Gate phase Scintigraphy"). By computer evaluation of this phase scintigraphy, we can then assess the pulsation of the walls of the ventricles and atria and create volumetric curves during the cardiac cycle, from which we can determine a number of quantitative parameters of systolic-diastolic heart activity.
Radiopharmaceuticals:
A radioindicator should be used that is maintained at a sufficient period of time a stable concentration and does not leak from the bloodstream. They are
99mTc-labeled erythrocytes, which can be prepared in two ways: 1. Laboratory in vitro from a blood sample, wherein the labeled erythrocytes 99mTc is reinjected back to the patient. 2. In vivo, in which the patient is first administered the dissolved tin salt and after about 15 minutes the required 99mTc-pertechnetate activity is administered. Tin ions Sn2+ will allow the binding of technetium to red blood cells in the circulation.
Execution:
After application of approx. 400 MBq of radiopharmaceutical is scanned by a camera detector, aimed at the heart area in the LAO projection at an angle of about 35-50°, so that the ventricular septum is approximately perpendicular to the plane of the detector. The patient has attached ECG electrodes (cardiomonitor), the R-wave output of which is connected to the camera's synchronization circuit. Let's wait for the heart rate to stabilize. We acquire in the ECG-gated mode so that the heart cycle is divided into about 16-32 images, reserved in the memory of the acquisition computer. We record about 500-800 heart cycles, while cycles with premature or delayed R-wave are eliminated. We can perform the examination at rest and in ergometric or pharmacological load.
Evaluation:
In the simplest case, the regional motility of the left ventricular wall can be assessed visually by means of cinematographic projection of individual images of a representative cardiac cycle in rapid succession (which makes visible the movement of the heart sections - "heartbeat"). It can be assessed semiquantitatively using the contour method, where the evaluation program plots the contour of the chamber in the end-diastole ED and end-systole ES into a single image. On the mutual relation of these contours we can recognize disorders - hypokinesia, akinesia, dyskinesia
( Fig.3.1.2 in "Radionuclide ventriculography"). Furthermore, we can construct parametric images distribution of a certain parameter in the organ. The simplest is the heart stroke image (image of pulse volume - difference of ED-ES images) and paradox image created by subtraction of ES-ED images, on which the atria are physiologically imaged and pathologically dyskinetic areas of the ventricle (which do not empty in systole but, on the contrary, increase). Furthermore, it is possible to compile an image of the ejection fraction, resulting from the image of heart rate by dividing by the image of the ventricular end-diastole (ED-ES)/ED. The most accurate analysis of cardiac cycle dynamics is provided by regional Fourier analysis using sine and cosine functions with certain amplitudes and phases in each pixel. We obtain two parametric images: the amplitude image, each site of which is proportional to the intensity of pulsation (local heart rate volume) and the phase image, expressing the time-phase shift (delay) of the onset of myocardial contraction at a given location, compared to the arrival of the ECG R-wave - Fig.3.1.3 , 3.1.4 in "Radionuclide ventriculography".
    On the ED and ES images of the representative cycle, we mark (manually or with the help of mathematical algorithms, including taking into account parametric images) the areas of interest (ROI) of the left ventricle, or even right ventricle and area of tissue background. The computer program creates a chamber volume curve from which important hemodynamic parameters are calculated: ejection fraction, heart rate, cardiac output, end-diastolic and residual volume of the ventricle, ejection and filling velocities of the ventricle - see pictures below. These parameters are especially important for the left ventricle; for the right ventricle, which has a less regular shape, accurate determination is more difficult.
    Mathematical analysis and computer evaluation of radionuclide ventriculography is described in detail in §3.1 "Radionuclide ventriculography" of the book "OSTNUCLINE".

Mathematical analysis and complex evaluation of radionuclide ventriculography
Evaluation:
On phase scintigraphic images of the cardiac cycle, nor on Fourier images of phase and amplitude, we do not observe regional heart wall motility disorders.

Conclusion:
Visual evaluation of images of individual phases of the cardiac cycle and quantitative analysis of cardiac dynamics indicate good global and local contractility of the walls of the left ventricle.
 
                                           
Signature: MUDr. Jozef Kubinyi, Ph.D.

Here are examples of the evaluation of normal á and heavily pathological â radionuclide ventriculography.

Mathematical analysis and complex evaluation of radionuclide ventriculography
Evaluation:
On phase scintigraphic images of the cardiac cycle and on Fourier images of phase and amplitude we observe the following regional left ventricular wall motility disorders :

   hypokinesia of segment :  Virtually all except posterolat.
        
akinesia of segment :
    asynchrony of segment :
     dyskinesia of segment : Apical  - very extensive !  

Conclusion:
Visual evaluation of images of individual phases of the cardiac cycle and quantitative analysis of cardiac dynamics indicate a severe disorder of left ventricular wall contractility with extensive hemodynamically significant apical dyskinesia. Extremely reduced ejection fraction of dilated LV.
                                                                                                                                              Signature: MUDr. Jozef Kubinyi, Ph.D.

Dynamic Bolus Angiocardiography
Purpose:
Rapid dynamic scintigraphy of the passage and dilution of a radioactive bolus through the right heart, lungs, and left heart, which provides information about the dynamics of blood flow in cardiac structures and large vessels. By analyzing this dynamics, we can obtain quantitative parameters of chamber function, their volume parameters, detection and quantification of intracardiac shunts. Together with radionuclide ventriculography, the regurgitation fraction of the left ventricle can be determined.

Radiopharmaceuticals:
Ordinary
99mTc-pertechnetate or better 99mTc-DTPA can be used for dynamic angiocardiography alone. If gated ventriculography is subsequently performed (eg to quantify regurgitation), 99mTc-labeled erythrocytes must be used as described above for ventriculography.
Execution:

Application of tracer (about 400 to 800 MBq) are done under a camera, directed at the area of the heart and lungs in the right oblique slope detector cameras 30-45°
(in this projection, the images of the ventricles, atria, lungs and aorta are optimally separated and distinguished). It is applied to the shortest possible vascular distance to the heart - to the anecubital vein of the right hand, to the subclavian vein or to the internal jugularis. Since this is a dynamic scintigraphy of fast process, it is necessary to perform a so-called bolus application (lat. bolus = bite ): fast single application of a radioindicator with high volume activity in a small volume of approx. 0.5 ml., in a short time approx. 1 s. (with rinsing of several ml of physiological solution using a three-way valve), with immediate start of dynamic scintigraphy with a sufficiently high frame rate (approx. 4 frames/sec.). We scan the fast phase for approx. 60-100 sec., then a slower sensing of the equilibrium phase can follow (10 sec./frame for approx. 5 min.).
Evaluation:
By visual inspection of sequential images of the passage of the bolus through the cardiac circulation, we can qualitatively assess possible abnormalities, especially premature circulation and recirculation, which could be caused by heart shunts. Then we mark the regions of interest (ROI) and create curves distribution of the radio indicator from the right and left ventricles, right atrium, lungs and possibly aorta.
    To detect and quantify the LR shunt, we use the time course of radioactivity in the lungs - pulmogram. Under normal circumstances, on this pulmogram, in addition to the sharp peak of the first flow, after about 30-50 sec. appears only a low broad peak of systemic recirculation, caused by the return of the radioindicator through the systemic circulation back to the heart. However, if an LP short circuit is present, another premature recirculation peak will appear soon after the peak of the first flow (or in the case of a small shunt, only the extension of the descending arm of the curve), caused by recirculation of shunt-passed blood from the left ventricle to the right ventricle and then to the lungs. By mathematically decomposing the pulmogram into the curve of the first pass, system recirculation and short-circuit recirculation, we can quantify the magnitude of the shunt using the ratio of short-circuit flow and lung flow Q
z/Qp (without shuntt is close to 0) or ratio of lung flow and system flow Qp/Qs (without shunt is approaching 1) - for details see "Bolus radiocardiography", Figures 3.2.3 and 3.2.4. From the right and left ventricular curves, we can determine the mean transit time of the central circulation and the cardiopulmonary blood volume. If a peak of premature circulation is present on the aortic curve, the right-left shunt can be detected and quantified by decomposition of the curves.
    If the equilibrium phase was sensed, by analyzing the curve from the left ventricle, it is possible
(a combination of the dilution and Stewart-Hamilton principles - see "Bolus radiocardiography" fig. 3.2.1 and 3.2.2 ) to detrmine the cardiac output. By analyzing the curve from the right ventricle, the ejection fraction of the right ventricle can in principle be determined. The combination of bolus angiocardiography and gated ventriculography makes it possible to quantify regurgitation in heart valves - to determine the regurgitation fraction.
  Mathematical analysis and computer evaluation of radionuclide angiocardiography is described in detail in §3.2 "
Bolus radiocardiography" of the book "OSTNUCLINE".

Mathematical analysis and complex evaluation of radionuclide angiocardiography
Visual assessment:
After intrajugular bolus injection of the radiolabel, the unmagnified cavities of the right heart are imaged, followed by the filling of the unexpanded lung artery and pulmonary circulation, which empties reasonably rapidly into the normally configured cavities of the left heart and aorta. During the passage of the bolus through the left heart, we do not observe a premature occurrence of a radioindicator in the right heart and lungs.

Conclusion:
In the visual evaluation of sequential scintigrams of the passage of the bolus through the cardiac circulation, nor in the quantitative analysis of circulation curves, we do not observe pathological changes in central hemodynamics.
  
                                                                          Signature: MUDr. Jozef Kubinyi, Ph.D.  

Here is the final protocol for the evaluation of angiocardiography without a short circuit á (but with regurgitation) and the intermediate results in the evaluation of a patient with a marked left-right shunt â .

Mathematical analysis of LP shunt in the evaluation of radionuclide angiocardiography
By decomposing the pulmogram curve into primary circulation, systemic recirculation and premature recirculation, we observe a hemodynamically significant left-right shunt . A comparison of the bolus passage curves through the right atrium and ventricle indicates this shunt at the level of the ventricular septum.

Myocardial perfusion scintigraphy
Purpose:
It is a non-invasive method for the assessment of regional myocardial perfusion and the effect of coronary stenosis for blood supply to the heart muscle in the relevant basin - at rest and during physical or pharmacological stress. It can be used to detect ischemic heart disease, its location, extent and degree of myocardial damage, assessment of myocardial cell viability.
Radiopharmaceuticals:
The basic requirement is that the radioindicator is efficiently taken up by myocardial cells already at the first flow and fixed in them (without redistribution) for the duration of scintigraphy. In this case, the displayed distribution of the radiopharmaceutical in the heart muscle is proportional to the "supply" - the regional flow in the coronary artery. However, this distribution also depends on the functional state of the cardiomyocytes: the accumulation of the radiolabel does not occur in areas where the heart cells are necrotic (or are replaced by connective tissue after infarction). Thus, the distribution of the radiolabel in the myocardium is proportional to the regional blood flow through the myocardium and the viability of myocardial cells.
  Previously, mainly thallium,
201 Tl- chloride, was used, as an analogue of potassium. It enters myocytes via the cell membrane mostly through the active process of the Na/K ATP (adenosine triphosphate) system, partly also by passive diffusion. The disadvantage of thallium is the low energy of X and gamma radiation, which causes poorer resolution and significant absorption of radiation in the tissue; also a considerable radiation load (to which abundant Auger electrons also contribute). Approx. 100 MBq 201Tl is applied.
Note: In the near future, however, a partial "renaissance" of thallium can be expected in connection with the introduction of special semiconductor CZK cameras in nuclear cardiology. These gamma cameras have a higher detection efficiency for low-energy photon radiation of about 73keV 201
Tl.
  Currently, 99m Tc - labeled radioindicators are mainly used for scintigraphy of myocardial perfusion and assessment of its viability, which provide significantly higher quality images at lower radiation exposure. They are mainly 99mTc-isonitriles - non-polar lipophilic complexes that enter myocardial cells by passive transport and bind in their cytoplasm or mitochondria. The radiopharmaceutical accumulates, depending on the blood supply, in healthy viable cells, while in cells damaged (eg due to ischemia) or even dead and replaced by scar fibrous tissue, no accumulation occurs. The distribution of the radioindicator at individual sites of the myocardium is then proportional to the regional blood flow through the myocardium and the viability of myocardial cells.
  The concentration of these
99mTc-isonitrile radiopharmaceuticals in myocytes remains stable for several hours after i.v. administration and thus shows the immediate perfusion situation of the myocardium at the time of application, eg during exercise. The most common perfusion radiopharmaceuticals are 99mTc-MIBI (2-methoxyisobutyl-isonitrile) and 99mTc-Tetrofosmin (diphosphine complex). Approx. 500-800 MBq of 99mTc MIBI or tetrofosmin is applied.
PET radiopharmaceuticals for perfusion myocardial scintigraphy 
Myocardial scintigraphy using positron emission tomography (PET) is performed relatively infrequently. On the one hand, for more demanding PET instrumentation, but mainly due to the difficult availability of suitable positron radionuclides and radiopharmaceuticals (discussed above in §4.8., Section "Radionuclides and radiopharmaceuticals for PET").
  The simplest PET radioindicator for perfusion is "labeled water", in which ordinary oxygen
16O is replaced by a positron radionuclide 15O - ie water H215O. After application, it passes through free diffusion through capillaries and cell membranes, so that the distribution of radioactivity, measured by PET, is proportional to blood flow. However, due to the high concentration of radioindicators in the bloodstream, myocardial imaging is not very contrasting. Another perfusion radioindicator is ammonia 13NH3 labeled with nitrogen-13. Thanks to the high first extraction (80%) and linear uptake according to blood flow in the myocardium, it provides quality images. PET scintigraphy with these short-lived radionuclides 15O (T1/2 = 2min.) and 13N (T1/2 = 10min.) is bound to centers with a cyclotron and is used mainly for research purposes.
  The short-term positron radionuclide rubidium
82 Rb is sometimes used for scintigraphy of myocardial perfusion by the PET method due to the fact, that it can be obtained from 82Sr/82Rb generator in the workplace. It is applied in the form of chloride 82RbCl, behaving as an analogue of potassium (similar to thallium 201 Tl mentioned above; compared to thallium, however, rubidium-82 provides better quality scintigraphic images at lower radiation exposure).
    The most common PET radiopharmaceutical fluoro-deoxyglucose
18FDG, although it is mainly used for tumor imaging, but as a glucose analogue, it is also taken up in the myocardium depending on perfusion, ischemia and viability (see "Metabolic scintigraphy, myocardial viability imaging" below).
Execution:
The examination can be performed at rest or under load. At rest conditions, the distribution of blood flow in the myocardium is usually homogeneous, minor or moderate perfusion disorders do not manifest. If the coronary stenosis is not greater than about 90%, the resting flow through the myocardium is sufficient to ensure normal myocardial metabolism; we get a normal perfusion scintigram of the myocardium, despite possibly presence of ischemic heart disease. Disorder of myocardial perfusion, even without exercise, can only be observed in severe coronary artery disease or after myocardial infarction.
Diagnostic sensitivity to perfusion disorders will only become apparent during examination under stress, when the requirements for oxygen supply to the heart tissue will increase - for increase coronary blood flow. Normal coronary arteries respond to this by visodilation and a corresponding increase in coronary flow, which is reflected in an increased concentration of radioindicator in perfusion scintigraphy. However, pathologically narrowed coronary vessels are not capable of this
(according to their possibilities, they are already dilated by compensatory mechanisms even at rest), the load has only a small effect on their flow. In the basin of the coronary artery, which is affected by hemodynamically significant stenosis, relatively lower perfusion is manifested during stress than in the surrounding parts of the myocardium - on the scintigram this place appears as a perfusion defect in the myocardium, or at least as a reduction in radiopharmaceutical distribution.
Therefore, i.v. application of the radiopharmaceutical is performed during exercise, either physical (usually a bicycle ergometer) or pharmacological - application of vasodilators (dipyridamole, adenosine, dobutamine).
    We perform our own scintigraphy in about 10 minutes after application. Scintigraphic scanning was previously performed planarly
(in LAO projection 30°, 60°) before the introduction of SPECT tomographic scintigraphy, but SPECT tomographic scintigraphy provides better differentiation of individual parts of the myocardium (advantages of tomographic scintigraphy over planar were discussed above in §4.3 "Tomographic scintigraphy"). The most commonly used is a two-detector SPECT camera *), whose detectors are set to an angle of 90°. Approx. 32-64 projections are recorded at a total angle of 180° around the patient - from the right front oblique projection of RAO 45° to the left rear oblique projection of LPO 45°. An ECG-gated myocardial SPECT is performed with R-wave synchronization of the electrocardiogram, analogously as described above for gated ventriculography, we captures about 500 cycles.
*) Single-purpose special types of cardiological cameras optimized for myocardial scintigraphy have also been developed. At some workplaces, perspective semiconductor CZK cameras are beginning to be used for SPECT myocardium
(§4.2., Part "New and alternative physical and technical principles of scintillation cameras", passage "Semiconductor multidetector cameras", Fig.4.2.10 on the right). When using them, it is sufficient to apply less than half of the usual activity, while reducing the examination time.
    
If stress perfusion scintigraphy is normal, resting scintigraphy is no longer necessary.
Evaluation:
On tomographic images of coronary, sagittal and transverse sections, the myocardium is displayed *) in the form of more or less closed "rings" or "rolls" or "horseshoes", on which we can visually evaluate the distribution of radioactivity - assess the size, number and location of perfusion defects and hypoperfusions.

* ) In the pictures we can clearly see only the thicker wall of the left ventricle, not the thinner wall of the right ventricle.

    For the semiquantitative evaluation is often used structure so-called polar maps
(slang "bull's eye"): coronal slices perpendicular to the short axis of the left ventricle are transform and sum to each other to form concentric circular profile with a tip in the center. The brightness (or color) of the resulting circular area is modulated by the different concentration of the radioindicator shown in the summation of the individual layers. These profiles are stored as concentric rings in a new circular image. This gives a clear normalized polar map of the distribution of activity in the myocardial wall - a map of regional perfusion of the myocardium (in polar coordinates centered in the apex), which is divided into segments corresponding to the basin of individual coronary arteries (according to international recommendations, 17 segments are used). The relative perfusion values in the individual segments are compared with the corresponding values in normal patients stored in the normal database. This results in relative indices - the so-called scores, which help determine the severity of the myocardial perfusion disorder (and possibly the corresponding risk).
    Visual images and polar maps are evaluated in scintigraphy under stress and at rest. In the case of ECG-gated SPECT myocardium, in addition to perfusion, we can also determine the ejection fraction of the left ventricle.

Evaluation of stress + rest scintigraphy of
myocardial perfusion with 4DM SPECT program.

We observe a significant reduction in perfusion in the
basal segment of the lower wall, partially
reversible (results from a comparison of polar maps
at rest and stress.

(scintigraphic images were taken by
MD. M.Havel, PhD., KNM FN Ostrava)

Metabolic scintigraphy, imaging of myocardial viability
Perfusion scintigraphy alone (showing hypoperfusion sites), or analysis of general and regional contractile systolic function (or dysfunction), may not be able to assess the maintenance of myocardial tissue viability, which may be temporarily hibernated due to hypoperfusion. To assess the function and condition of the myocardial tissue itself, myocardial cells, it may be useful to visualize metabolism glucose, fatty acids or amino acids. Appropriate biochemically active compounds labeled with radionuclides, especially positron isotopes of biogenic elements, are used for this purpose. These imaging methods have the ability to independently (separately or simultaneously) assess blood flow and metabolism. Areas of myocardium with reduced flow, but preserved metabolism can thus be recognized - revascularization may be useful here.
    In clinical practice, PET examination of myocardial glucose metabolism using
18F-FDG (fluorodexyglucose) is sometime used. Under normal circumstances of a well-perfused viable myocardium, myocytes gain the necessary energy mainly by beta-oxidation of fatty acids. However, in ischemia, anerobic glycolysis is becoming more important as a source of energy for the work of cardiac cells. After i.v. administration, 18F-FDG accumulates to an increased extent in the ischemic, but viable, areas of the myocardium, while it does not accumulate in the normally perfused areas (because glucose is not used to obtain energy there). FDG also does not accumulate in the non-viable or necrotic myocardium because metabolism does not take place there at all. By comparing between regional blood flow and metabolism, we get information about normal, hibernating and necrotic myocardium. In the case of ischemic defects with preserved viability, it is useful to perform revascularization (by-pass or angioplastiy, stent), after which there is a "revival" of these sites and often an improvement in myocardial contractility and an increase in overall cardiac performance.
  In particular for research and experimental studies, imaging of the distribution of free labeled fatty acids such as
123I- labeled BMIPP penta- and hexadecanoic acid derivatives for SPECT is used to assess metabolism. For PET, it is 11C-palmitate for fatty acid metabolism and 11C-acetate (which is incorporated into the Krebs cycle) for oxidative metabolism.....
Receptor scintigraphy of the myocardium 
The spread of excitations in the myocardium and the regulation of coronary flow are determined by the function of the sympathetic nervous system in the heart. This function depends, among other things, on the distribution of receptors, especially noradrenaline receptors. The distribution of these receptors can be visualized by
123 I -labeled MIBG, which binds to them. ...........

Myocardial amyloidosis scintigraphy
Amyloidosis *) is a systemic disease, caused by a conformational change of the geometric structure of some originally soluble proteins into an insoluble form of amyloid fibrils, which can be deposited in the interstitial areas of various tissues and organs - brain, heart, kidneys.... This can adversely affect their structure and physiological functions.
*) This misleading name (Greek amylon = starch) arose from the fact that it behaved similarly to starch when stained with Lugol's solution. Later, however, biochemists showed that amyloid has the chemical composition of proteins.
  Cardiac amyloidosis is caused by the deposition of insoluble proteins in the form of amyloid fibrils in the intercellular spaces of the heart myocardium. It leads to thickening of the heart wall and deterioration of its contractility. The most common types of cardiac amyloidosis are transthyretin (TTR) and light chain (AL) protein amyloidosis.
Purpose :
It is a non-invasive method for revealing the presence and severity of deposition of insoluble amyloid proteins in the myocardium, which could be the cause of deterioration of heart function (cardiomyopathy). It makes it possible to distinguish TTR amyolidosis from other causes of thickening of the heart chamber walls
(seen on echocardiography).
Radiopharmaceuticals :
In planar and SPECT scintigraphy, osteotropic radiopharmaceuticals such as
99mTc-DPD (99mTc-3,3-diphosphono-1,2-propanedicarboxylic acid), 99mTc-pyrophosphate, or 99mTc-HMDP (hydroxy-methylene-diphosphonate) are used here - see §4.9.7 "Scintigraphy of the skeleton". These radiopharmaceuticals are accumulated not only in the skeleton, but also in the extracellular space in the myocardium, filled with fibrils of amyloid proteins.
  Radioindicators 18F-florbetapir, 18F-florbetaben and 18F-flutemetamol are used in PET.
Execution :
After i.v. application of approx. 200 MBq radioindicator, static scintigraphy of the chest region in AP projection is performed in approx. 2 hours. It can also be part of whole-body scintigraphy of the skeleton
(cardiac amyolidosis can sometimes appear even as a secondary finding in classic whole-body scintigraphy of the skeleton). For a more detailed assessment of the anatomical correlation, it is also advisable to obtain a SPECT/CT scintigraphy of the chest.
Evaluation :
On scintigraphic images of the chest, we assess how differently the radiopharmaceutical accumulated in the skeleton - in the rib bones - and in the myocardium. The presence and severity of cardiac amyloidosis is often assessed here using an empirical so-called Perugini score :
Perugini score 0: The osteotropic radiopharmaceutical was captured only in the skeleton, the heart myocardium is not shown. Scintigraphy here is not indicative of cardiac amyloidosis.
Perugini score 1: Uptake of the radiopharmaceutical in the myocardium is present, but less than in bone. It is a suspected finding of AL or TTR amyloidosis
(endomyocardial biopsy should be used to confirm or refute).
Perugini score 2: The same or greater uptake of the radiopharmaceutical is observed in the myocardium than in the skeleton. It is an indication for TTR-amyolidosis.
Perugini score 3: Significant uptake of the osteotropic radiopharmaceutical in the thickened myocardium is present, with greatly weakened or absent uptake in the skeleton. It is highly specific for TTR-amyolidosis.
  Here are examples of chest scintigraphic images for different amyloidosis scores
(they are older, lower quality planar images...) :

PET scintigraphy has not yet shown significant progress over planar/SPECT scintigraphy. However, some results show with PET with 18F-florbetaben that images obtained a few minutes after application could help distinguish AL from TTR amyolidosis..?..


4.9.5. Lung scintigraphy (nuclear pneumology)
The lungs
(Latin pulmo , Greek pneumo ) are a key respiratory organ in higher animals and humans. They exchange gases - especially oxygen and carbon dioxide *), between blood and air. The chain [ventilation of the pulmonary alveoli ® diffusion of gases through the alveolocapillary membranes ® perfusion of the lungs ® blood circulation] transports oxygen from the air to the cells of tissues and organs and removes carbon dioxide from the tissues into the atmosphere.
*) The main source of energy for cells during metabolism is oxidation (especially of glucose), which produces "energetic" molecules such as ATP (adenosine triphosphate), the "waste" products are mainly water and carbon dioxide. Thus, gas exchange is required for the metabolism to function - oxygen supply and carbon dioxide removal. In higher organisms, this gas exchange is not sufficient by passive diffusion, but takes place by active breathing (respiration, ventilation) of air from the atmosphere through the lungs (in fish by oxygen exchange from the water in the gills).
    The lungs have a spongy consistency, they consist of more than 300 million pulmonary cellars - alveoli, which are small hollow thin-walled sacs (about 150
mm in diameter). Their wall - the alveolocapillary membrane - is formed by one layer of thin cells, type I pneumocytes (wall thickness is about 1 mm). The total area of the alveolocapillary membrane is about 60 m2. There are also thicker type II pneumocytes in the alveolar wall, which produce substances that reduce surface tension (surfactant) and macrophages, which phagocytose dust and foreign particles.
  Respiratory gases diffuse through the membrane of the alveoli in the direction of pressure and concentration gradients; depends on the partial pressure of these gases in the inhaled air and in the non-oxygenated blood flowing in the capillaries around the alveoli. Oxygen has a lower partial pressure in deoxygenated blood and therefore passes through the membrane from the alveoli to the blood. Carbon dioxide, on the other hand, has a higher partial pressure in the venous blood and therefore passes from the capillaries through the membrane into the air in the alveoli, from where it is exhaled.
  Anatomically, the lungs are a pair organ - the left and right lungs. They are divided into lobes (3 lobes have the right lung, 2 lobes have the left). The lobes are further divided into bronchopulmonary segments, each of which has its own air and blood supply.
    Air is fed into the lungs (and removed) trough bronchi, which are tubes with cartilaginous walls,
that branch many times inside the lungs into even finer tubes, down to the alveoli. Airflow in the lungs - respiration - takes place alternately inhalation (inspirium) and exhalation (expirium). When inhaling, the contraction of the intercostal muscles and the diaphragm increases the volume of the thoracic cavity, and new air is drawn into the lungs by the airways due to the negative pressure. During exhalation, the air used is blown out of the airways into the atmosphere by the passive pressure when the chest is contracted.
    Blood circulation of the lungs - pulmonary perfusion - starts the pulmonary artery, emanating from the right ventricle and delivering non-oxygenated blood. In the lungs, they branch many times up to the capillaries that surround the alveoli. Here, oxygen diffuses into the blood and carbon dioxide into the alveoli. The vessels carrying the oxygenated blood connect in the pulmonary veins, which open into the left atrium. From there, the left ventricle pushes oxygenated blood through the aorta into the bloodstream, distributing it throughout the body. Deoxygenated blood, which is also enriched with carbon dioxide, then leads through the venous system to the right atrium, from where the right ventricle returns it to the lungs for further oxygenation and CO
2 removal - small cardiac circulation ( pulmonary, cardiopulmonary), continuing to the left atrium. Oxygenated blood is then pumped through the left ventricle into the large blood circulation - systemic. Normally, almost the same amount of blood flows through the pulmonary circulation as the systemic circulation, but under lower pressure.
Pathology of the lungs and respiratory system : 
Pulmonary edema ("emphysema"), as a result of pulmonary hypertzenze, increased pressure in the pulmonary circulation, left ventricular failure, ....
Pulmonary embolism
is a blockage of the pulmonary vessels from the venous system or the right heart due to thrombosis. It leads to hypoperfusion or aperfusion of certain parts of the lungs. Sudden obstruction of the middle lung branches is called a pulmonary infarction.
Bronchial asthma
- spasm and bronchial constriction associated with dyspnea ........
Pneumoconiosis  is a ventilation disorder caused by fibrosis due to prolonged inhalation of dust, such as silica (silicosis), popularly called "dusting of the lungs". .......
Inflammatory and infectious diseases
, tuberculosis.....
Tumors
- primary lung cancer, metastases of other tumors to the lungs....
  
    Three basic conditions must be met for the order functioning of the respiratory system: 1. Good pulmonary perfusion; 2. Good ventilation of the pulmonary alveoli; 3. Proper function of the alveolar membrane. Methods of nuclear medicine also focus on the diagnosis of these components of the respiratory system.  

Pulmonary perfusion scintigraphy
Purpose:

Imaging the distribution of capillary perfusion in the lung parenchyma, revealing regional perfusion defects due to embolization or other involvement of the pulmonary arterial river system (compression by inflammatory or tumor foci, pleural effusion, .....).
Radiopharmaceuticals:
99mTc-MAA (labeled macroaggregate of albumin, or albumin microspheres).
Execution:

After iv application of approx. 200 MBq of radioindicator, static scintigraphy of the lung area is performed in 4-6 projections, the most important of which are PA and AP projections.