Добавил:
Sekretar
kiopkiopkiop18@yandex.ru
t.me/Prokururor I Вовсе не секретарь, но почту проверяю
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз:
Предмет:
Файл:Ординатура / Хирургия / Библиотека им академика М.И. Перельмана / Книга_5255_Библиотеки_им_академика_М_И_Перельмана
.pdf
Combined PET/CT:Technical aspects and image registration322
radionuclide data (10). Furthermore, virtually all clinical radionuclide imaging detectors use
PMTs whose performance can be seriously affected in the presence of magnetic fields. This is
especially problematic in an MRI scanner, which relies on rapidly switching gradient magnetic
fields and radiofrequency (RF) signals to produce the magnetic resonance image. The presence
of the magnetic field gradients and RF signals certainly could disrupt the performance of a PMT
and PET detector if they were located within or adjacent to the magnet of the MRI system.
Similarly, the operation of the MRI system relies on a very uniform and stable magnetic field to
produce the MR image. The introduction of radiation detectors, electronics, and other bulk
materials can perturb the magnetic field in a way that introduces artefacts in the MR image.
In spite of these challenges, several research groups are investigating methods to integrate a
PET system directly in an MRI scanner by designing detectors made from nonmagnetic materials
that can be placed within the magnetic field of an MRI/MRS system (21). For example, the
UCLA group developed a 3.8-cm ring of small scintillator crystals that was placed in the MR
system for PET imaging (35,39). The crystals were optically coupled through 3 m long fibre
optics to an external array of position-sensitive photomultiplier tubes, and which could be readout through external processing electronics. By keeping the radiation-sensitive elements of the
detector within the MR system, while operating the electronics away from the magnetic field,
the combined system could perform simultaneous PET/MR imaging. The same group also
performed simultaneous PET/MR imaging with a larger (5.6 cm-diameter) detector ring using
the same design (35,36). Their collaborators at Kings College London placed the system inside
of a 9.4-T NMR spectrometer to study metabolism in an isolated, perfused rat heart model. 32PNMR spectra were acquired simultaneously with PET images of 18F-FDG uptake in the
myocardium (40,41). This design concept is being extended to develop an MR-compatible PET
scanner with one ring of 480 LSO crystals arranged in 3 layers (160 crystals per layer) with a
diameter of 11.2 cm corresponding to a 5 cm diameter field of view, large enough to accommodate
an animal within a stereotactic frame (39). The system is designed to offer adequate energy
resolution and sensitivity for simultaneous PET/MRI imaging of small animals.
Other investigators have proposed PET/MRI systems configured with suitable solid-state
detectors that can be operated within a magnetic field for radionuclide imaging. Some groups
have tested avalanche photo diodes (APDs) within a high-field (7-T) NMR spectrometer and
have produced radionuclide data that appear to be free of distortion (42). However, it is still
unknown whether the introduction of the APD’s cause distortions in the magnetic field to an
extent that would cause severe artefacts in the MR image (21).
In this chapter, a brief overview of current state-of-the art developments of combined PET/
CT instrumentation is presented. We emphasize that many different design paths have been and
continue to be pursued in both academic and corporate settings that offer different trade-offs in
terms of their performance. It is still uncertain which designs will be incorporated into future
clinical systems, but it is certain that technological advances will continue and will enable new
quantitative capabilities in clinical oncologic imaging. Reliable molecular imaging plays a valuable
role in the assessment of cellular targets, evaluation of response to therapy, differential diagnosis,
prediction or selection of patients who will benefit from treatment, and in dosimetry for targeted
Combined PET/CT:Technical aspects and image registration 323
therapy. PET is poised to advance the application of molecular diagnosis in oncology, neurology,
cardiology, infectious diseases, and other types of disease. Nevertheless, PET/CT is obviously
not the only major non-invasive tool for the assessment of human disease. Major new technologies,
such as high-field MRI, bioluminescent and fluorescent imaging, and many other technologies,
have now blurred the artificial distinction that once set nuclear medicine apart as a “functional”
rather than “anatomic” imaging modality (43). Nonetheless, PET/CT maintains an exclusive
standing in the delivery of targeted therapies, but its superior picomolar sensitivity is being
challenged by competing technologies such as those using ultra small super paramagnetic contrast
agents (44).
References
1. Hasegawa B, Zaidi H. Dual-modality imaging: more than the sum of its components. In Quantitative
analysis in nuclear medicine imaging. H. Zaidi, Ed. Springer, New York, 35-81: 2005.
2. Hasegawa BH, Iwata K, Wong KH, et al. Dual-modality imaging of function and physiology. Acad
Radiol 2002; 9: 1305-1321.
3. Beyer T, Townsend D, Brun T, et al. A combined PET/CT scanner for clinical oncology. J Nucl Med
2000; 41: 1369-1290.
4. Townsend DW, Carney JPJ, Yap JT, Hall NC. PET/CT today and tomorrow. J Nucl Med 2004; 45: 4S-
14.
5. Wahl RL. Why nearly all PET of abdominal and pelvic cancers will be performed as PET/CT. J Nucl
Med 2004; 45: 82S-95S.
6. Goerres GW, Schulthess GK von, Steinert HC. Why most PET of lung and head-and-neck cancer will
be PET/CT. J Nucl Med 2004; 45: 66S-71S.
7. Cohade C, Wahl RL. Applications of positron emission tomography/computed tomography image
fusion in clinical positron emission tomography - clinical use, interpretation, methods, diagnostic
improvements. Semin Nucl Med 2003; 33: 228-23.
8. Israel O, Keidar Z, Iosilevsky G, et al. The fusion of anatomic and physiologic imaging in the
management of patients with cancer. Semin Nucl Med 2001; 31: 191-205.
9. Pfannenberg AC, Eschmann SM, Horger M, et al. Benefit of anatomical-functional image fusion in
the diagnostic work-up of neuroendocrine neoplasms. Eur J Nucl Med Mol Imaging 2003; 30: 835-
843.
10. Zaidi H, Hasegawa BH. Determination of the attenuation map in emission tomography. J Nucl Med
2003; 44: 291-315.
11. Ay M, Zaidi H. Assessment of errors caused by x-ray scatter and use of contrast medium when using
CT-based attenuation correction in PET. Eur J Nucl Med Mol Imaging 2006; 33: in press.
12. Ay M, Zaidi H Computed Tomography-based attenuation correction in neurological positron emission
tomography: evaluation of the effect of x-ray tube voltage on quantitative analysis. Nucl Med Commun
2006; 27: 339-346.
13. Zaidi H. Recent developments and future trends in nuclear medicine instrumentation. J Med Phys
2006; 16: 5-17
14. Schoder H, Larson SM, Yeung HWD. PET/CT in oncology: Integration into clinical management of
lymphoma, melanoma, and gastrointestinal malignancies. J Nucl Med 2004; 45: 72S-81S.
Combined PET/CT:Technical aspects and image registration324
15. Israel O, Mor M, Gaitini D, et al. Combined functional and structural evaluation of cancer patients
with a hybrid camera-based PET/CT system using F-18 FDG. J Nucl Med 2002; 43; 1129-1136.
16. Townsend DW, Beyer T. A combined PET/CT scanner: the path to true image fusion. Br J Radiol
2002; 75 Spec No: S24-30.
17. Townsend DW, Beyer T, Blodgett TM. PET/CT scanners: a hardware approach to image fusion.
Semin Nucl Med 2003; 33: 193-204.
18. Steinert HC, von Schulthess GK. Initial clinical experience using a new integrated in-line PET/CT
system. Br J Radiol 2002; 73: S36-S38.
19. Hasegawa BH, Brown JK, Kalki K, et al. SPECT-CT: Research and development. Medica Physica
1996; 12: 52-61.
20. Townsend DW, Beyer T, Kinahan PE, et al. The SMART scanner: a combined PET/CT tomograph for
clinical oncology, IEEE Nuclear Science Symposium and Medical Imaging Conference 1998
21. Townsend DW, Cherry SR. Combining anatomy and function: the path to true image fusion. Eur
Radiol 2001; 11: 1968-1974.
22. Ratib O. PET/CT image navigation and communication. J Nucl Med 2004; 45 Suppl 1: 46S-55S.
23. Blankespoor SC, Wu X, Kalki K, et al. Attenuation correction of SPECT using x-ray CT on an
emission-transmission CT system: Myocardial perfusion assessment. IEEE Trans Nucl Sci 1996; 43:
2263-2274.
24. Da Silva AJ, Tang HR, Wong KH, et al. Absolute quantification of regional myocardial uptake of
99mTc-sestamibi with SPECT: experimental validation in a porcine model. J Nucl Med 2001; 42:
772-779.
25. Bai C, Shao L, Da Silva AJ, Zhao Z. A generalized model for the conversion from CT numbers to
linear attenuation coefficients. IEEE Trans Nucl Sci 2003; 50: 1510-1515.
26. Kinahan PE, Hasegawa BH, Beyer T. X-ray-based attenuation correction for positron emission
tomography/computed tomography scanners. Semin Nucl Med 2003; 33: 166-179.
27. Nakamoto Y, Osman M, Cohade C, et al. PET/CT: comparison of quantitative tracer uptake between
germanium and CT transmission attenuation-corrected images. J Nucl Med 2002; 43:1137-1143.
28. Osman MM, Cohade C, Nakamoto Y, et al. Clinically significant inaccurate localization of lesions
with PET/CT: frequency in 300 patients. J Nucl Med 2003; 44: 240-243.
29. Townsend DW, Jakoby B, Hubner K, et al. Performance evaluation of a high sensitivity high resolution
PET scanner designed based on large area. [abstract] J Nucl Med 2005; 46: 206P.
30. Kastis GA, Furenlid LR, Wilson DW, et al. Compact CT/SPECT small-animal imaging system. IEEE
Trans Nucl Sci 2004; 51: 63-67.
31. Goertzen AL, Meadors AK, Silverman RW, Cherry SR. Simultaneous molecular and anatomical
imaging of the mouse in vivo. Phys Med Biol 2002; 21: 4315-4328.
32. Schramm NU, Ebel G, Engeland U, et al. High-resolution SPECT using multipinhole collimation.
IEEE Trans Nucl Sci 2003; 50: 315-320.
33. Beekman FJ, Vastenhouw B. Design and simulation of a high-resolution stationary SPECT system for
small animals. Phys Med Biol 2004; 49: 4579-4592.
34. Kiessling F, Greschus S, Lichy MP, et al. Volumetric computed tomography (VCT): a new technology
for noninvasive, high-resolution monitoring of tumor angiogenesis. Nat Med 2004; 10: 1133-1138.
35. Shao Y, Cherry SR, Farahani K, Meadors K. Simultaneous PET and MR imaging. Phys Med Biol
1997; 42: 1965-1970.
Combined PET/CT:Technical aspects and image registration 325
36. Slates R, Farahani K, Shao Y, et al: A study of artefacts in simultaneous PET and MR imaging using
a prototype MR comptatible PET scanner. Phys Med Biol 1999; 44: 2015-2027.
37. Marsden PK, Strul D, Keevil SF, Williams SC, Cash D. Simultaneous PET and NMR. Br J Radiol
2002; 75 (Spec No): S53-59
38. Yamamoto S, Takamatsu S, Murayama H, Minato K. A block detector for a multislice, depth-of-
interaction MR-compatible PET. IEEE Trans Nucl Sci 2005; 52: 33-37.
39. Slates R, Cherry SR, Boutefnouchet A, et al. Design of a small animal MR compatible PET scanner.
IEEE Trans Nucl Sci 1999; 46: 565-570.
40. Garlick PB. Simultaneous PET and NMR-Initial results from isolated, perfused rat hearts. Br J
Radiol 2002; 75 (Spec No): S60-66
41. Garlick PB, Marsden PK, Cave AC, et al. PET and NMR dual acquisition (PANDA): Applications to
isolated perfused rat hearts. NMR Biomed 1997; 10: 138-142.
42. Pichler BJ, Judenhofer MS, Catana C, Walton J, Kneilling M, et al. Performance test of an LSO-APD
detector in a 7-T MRI scanner for simultaneous PET/MRI. J Nucl Med 2006; 47: 639-647.
43. Park JM, Gambhir SS. Multimodality radionuclide, fluorescence, and bioluminescence small-animal
imaging. Proceedings of the IEEE 2005; 93: 771-783.
44. Delikatny EJ, Poptani H. MR techniques for in vivo molecular and cellular imaging. Radiol Clin
North Am 2005; 43: 205-220.
Image Correction Techniques in SPECT
and PET
Habib Zaidi
Nuclear medicine has a long tradition of incorporating quantitative analysis in its
diagnostic and therapeutic procedures. Many of the clinical and research applications of
molecular imaging rely on a solid quantitative foundation, which is highly dependent on the
performance characteristics of nuclear medicine instrumentation and the accuracy of image
correction and reconstruction algorithms. This text reflects the tremendous increase in
interest in the quantitative capabilities of standalone (SPECT and PET) and dual-modality
(SPECT/CT, PET/CT and PET/MR) molecular imaging as both clinical and research imaging
modalities in the pPast decade. New algorithmic developments of image correction and
reconstruction techniques are aiming at attaining better image quality and achieving more
accurate and automated quantification of physiological parameters of interest in clinical and
research settings. Impact of physical degrading factors including attenuation of photons and
contribution from photons scattered in the patient and partial volume effect on diagnostic
quality and quantitative accuracy of SPECT and PET data will be discussed. The fundamental
concepts of quantitative image analysis techniques as they are applied in diagnostic and
therapeutic nuclear medicine using dedicated SPECT and PET instrumentation and dualmodality imaging devices will be explored. Potential future applications of quantitative
molecular imaging are also addressed especially its use prior to therapy for dose distribution
modeling and patient-specific 3D dosimetry in treatment planning towards the concept of
image-guided radiation therapy.
Molecular imaging using single-photon emission computed tomography (SPECT) and positron
emission tomography (PET) has evolved into an academic field and is progressively gaining
importance in the clinical arena. Significant progress has been made in the design of highresolution SPECT cameras and PET units for whole-body clinical imaging and the development
of accurate quantitative imaging protocols incorporating accurate image correction techniques
and sophisticated reconstruction algorithms. However, emerging clinical and research applications
32 6
Image Correction Techniques in SPECT and PET 327
of molecular brain imaging promise even greater levels of accuracy and precision and
therefore impose more constraints with respect to the quantitative capability of these
technologies. Although the use of nuclear imaging for diagnosis and therapy has origins
dating back almost to the inception of nuclear medicine following the pioneering work of
G. de Hevesy, quantitative imaging has only recently emerged as a promising approach for
diagnosis and therapy of many diseases. Over this time, its use has evolved from a tool
available to only a few expert physicists and research biomedical scientists to one that may
be integrated in commercial software that is now potentially available for routine use. This
evolution has been driven by four fundamental and critical developments (1):
The ever-increasing sophistication of nuclear imaging instrumentation and associated
hardware developments for image correction (e.g. x-ray CT scanning for accurate
attenuation correction);
The rapid evolution and widespread clinical acceptance of PET and dual-modality
(SPECT/CT and PET/CT) imaging as effective diagnostic tools, requiring improved and
accurate quantitative analysis;
The availability of open source libraries for simulation, image registration, reconstruction,
processing and objective assessment of image quality methodologies; which spurred the
development of more complex and ambitious computational tools (e.g. anatomicallyguided reconstruction and partial volume correction);
The near exponential increase in preclinical research studies using small-animal imaging
devices; an area in which quantification is always essential.
It is worth emphasizing that quantification has been traditionally performed in the case of
PET, which started mainly as a research tool where there was greater emphasis on accurate
quantitative measurements, and more recently has been applied for SPECT. There are many
historical and methodological reasons for that, which have been addressed elsewhere (1). Different
strategies for image reconstruction and accurate attenuation, scatter and partial volume effect
corrections have been proposed so far with various degrees of success.
Figure 1 visualizes image quality degradation by comparing an ideal 18F-FDG PET study of
the Hoffman 3-D brain phantom without any physical or instrumentation-based limitation and a
realistic brain PET study obtained by Monte Carlo simulations including all degradation effects.
The differences in image quality between both images mainly consist of degraded spatial resolution
and contrast resolution, due to limited system resolution, contribution from scattered events, and
the limitations of the reconstruction algorithm. The difference is striking and show how
complicated it will be to properly obtain a correct activity quantification from PET imaging.
The number of scientific contributions related to this subject has been increasing steadily,
which spurred the writing of this invited contribution as a snapshot of the dynamically changing
field of quantitative radionuclide imaging. This chapter presents the physical and methodological
basis of quantitative image analysis and briefly summarizes state of the art developments in
algorithms aiming at accurate quantification of SPECT and PET data. Future prospects and
suggestions for future research will also be given.
GSPant\Newbook\Final-2008\21-chp\327
Image Correction Techniques in SPECT and PET328
True distribution
True distribution
Coincidence
DAQ-electronics
DAQ-electronics
PET camera
PET camera
Coincidence
Actual distribution
Actual distribution
Tracer
Tracer
511 keV
511 keV
511 keV
Annihilation process
Annihilation process
LOR
LOR
+
+
+
–
–
–
e
e
e
511 keV
511 keV
511 keV
Figure1: Principle of PET data acquisition showing expected differences in terms of image
quality and quantitative accuracy between the true activity distribution (left) and actual activity
distribution (right) for the Hoffman 3-D brain phantom. The differences are mainly due to intrinsic
limitations of the PET scanner and inherent imperfections of image correction and reconstruction
techniques.
Scatter Compensation Strategies
In the earliest literature on scatter correction, the main import of scatter was considered to be
a loss of contrast in the image. In the simplest of descriptions, this means that a true zero in a
reconstructed image occurs as a positive value. This effect was demonstrated by imaging nonradioactive spheres in a radioactivity surround. The corruption was frequently described as a
pedestal upon which the true image sat. It was soon after realized that for quantitative imaging,
Compton scatter causes a more complicated distortion in at least parts of the image. So, to the
extent that clinicians want an accurate quantitative image, including the best contrast possible,
scatter is always a problem. The extent to which it can be shown to have a disabling effect upon
the goal for which the image is to be employed is a much more difficult matter to discuss and to
document.
Scatter correction procedures conventionally used in planar imaging can be employed for
SPECT by correcting each projection separately, and vice versa.
A thorough review on SPECT and PET scatter correction methods as of 2004 is given by
Novikov (2). Scatter correction methods in SPECT can be divided into two broad categories:
those without explicit subtraction, and those employing subtraction. That division will not be
followed here, but methods will instead be divided into those that employ implicit correction and
those that use explicit correction. Implicit correction procedures are those, which require nothing
beyond satisfying some criterion. Explicit correction methods employ a procedure designed to
compensate for scattering. The compensation can be by putting the scatter counts back into the
voxel from which the gamma originated, by subtracting the scatter counts and then reconstructing,
or by reconstructing unscattered counts in such a way that the existence of scatter counts is
accounted for.
GSPant\Newbook\Final-2008\21-chp\328
Image Correction Techniques in SPECT and PET 329
Methods using a transmission measurement to provide additional, hopefully-pertinent
information about the scatter in a particular imaging situation are attractive and proved to work
well in many situations. Iterative reconstruction-based scatter compensation combined with
sophisticated Monte Carlo modelling approaches are particularly attractive from a theoretical
point of view and are documented in a few research papers. However, their clinical relevance and
applicability in a clinical environment still needs to be demonstrated.
Despite the importance of scatter for quantitative imaging, scatter correction strategies in
PET have only been briefly discussed in the literature (3) with the exception of extensive reviews
provided in book chapters (4). Figure 2 illustrates typical 18F-FDG thoracic PET scan reconstructed
with and without scatter correction. Scatter correction improves the contrast and removes artificial
uptake in the lungs. Therefore, there is consensus within the nuclear medicine community with
respect to the potential usefulness and necessity of scatter correction for either qualitative
interpretation of patient images or extraction of clinically-useful quantitative parameters.
Over the last two decades, many methods have been developed for the purpose of reducing
the resultant degradation of image contrast and loss of quantitative accuracy in PET due to
scattered events. The main difference among the correction methods is the way in which the
scatter component in the selected energy window is estimated. The most reliable method to
determine the actual amount of scatter in the image is accurate modeling of the scatter process to
resolve the observed energy spectrum into its unscattered and scattered components.
Figure 2. Example of clinical PET images reconstructed without (left) and with (right) scatter
correction for a typical 18F-FDG whole-body scan. Note that scatter correction improves the
contrast and removes artificial uptake in the lungs in myocardial imaging.
A number of scatter-correction algorithms for PET have been proposed in the literature.
They fall into four broad categories:
1. Multiple-energy-window (spectral-analytic) approaches;
2. Convolution/deconvolution-based approaches;
3. Approaches based on direct estimation of scatter distribution;
4. Statistical reconstruction-based scatter compensation approaches.
Different versions of the above methods have been successfully implemented for 3D PET
and are discussed by Zaidi (3). There is little in the literature reporting systematic studies on the
clinical impact of different scatter correction techniques versus no correction in 3D PET. It is
well known that subtraction-based scatter correction increases statistical noise. However, in
general scatter correction improves the contrast compared to the case where no correction is
GSPant\Newbook\Final-2008\21-chp\329
Image Correction Techniques in SPECT and PET330
applied. In particular, the low-count regions and structures are better recovered after scatter
compensation.
It can be argued that there is consensus within the nuclear medicine community with respect
to the potential usefulness and necessity of scatter correction for either qualitative interpretation
of patient images or extraction of clinically-useful quantitative parameters. The main application
which is still the subject of debate is 15O-[H2O] brain activation studies characterised by lowcount imaging protocols where scatter subtraction might jeopardize the power of statistical
analysis significance. In these cases, PET studies focus on identification of functional differences
between subjects scanned under different conditions. Whether the scatter component can be
considered as constant between the two conditions for inter-subject comparisons still needs to be
demonstrated. This constancy is required to confirm the hypothesis that the outcome of statistical
analysis (reflecting subtle changes in distribution of radiotracer) does not change greatly with
and without scatter compensation.
It is gratifying to see the progress that scatter correction has made in the last twenty years,
from very crude energy-based approaches, through analytic and Monte Carlo modelling, and
more recently iterative reconstruction-based scatter correction approaches. Recent developments
have been enormous, especially improved accuracy, precision, and computational speed, in
conjunction with decreased calibration data. The necessity for scatter correction is well understood
in research environments. Moreover, scatter correction in now carried out in some clinical
settings, even in institutions without extensive physics and computing support. Implementation
of validated techniques in commercial software packages would be useful to further attract the
interest of the clinical community. This greater interest would in turn lead to increased refinement
of scatter correction techniques. It is expected that with the availability of greater computing
power in the near future, more complex and ambitious computer-intensive scatter modelling and
correction algorithms will increasingly become clinically feasible (3).
Attenuation Correction Strategies
The accuracy achieved by attenuation correction procedures depends mainly on the rigor
followed to derive patient-specific attenuation map. Two broad classes have emerged: (i) calculated
(transmissionless) methods, which are based on an assumed anatomical model representing the
shape and spatial distribution of attenuation coefficients in the head and (ii) measured
(transmission-based) methods, which in general rely on supplementary acquisition of a
transmission scan. These techniques vary in complexity, accuracy, and computation time
required (5).
In the two-dimensional case, the fundamental relation that links the imaged object ƒ(x,y) and
corresponding attenuation map
central slice theorem. The general equation describing measured projections in term of the
radionuclide source distribution inside an attenuating medium is called the attenuated Radon
transform and is given in the case of SPECT by:
GSPant\Newbook\Final-2008\21-chp\330
to its 1D projection data
),( yx
is called the 2D
),(sp
Image Correction Techniques in SPECT and PET 331
yxl
sL
,
0,
','exp,,
drdlyxyxfsp
(1)
where l(x,y) is the distance from the emission point (x,y) in the object to the detector along
the line
is the angle between the rotating detector plane and the stationary
),,(sL
reconstruction plane and µ(x’,y’) the attenuation coefficient at position (x’,y’). Whereas, in
the case of PET, the attenuated Radon transform is given by:
sLsL
,,
,exp,,
drdlyxdryxfsp
(2)
The ideal solution would have been to use an exact solution for the inverse problem to solve
the Radon transform and reconstruct the spatial distribution of the tracer f(x,y). However, owing
to the complexity of the equation in the case of nonuniform attenuation, an exact solution is not
possible in general. The seminal contribution by Novikov (2) who recently gave an explicit
inversion formula for the attenuated Radon transform for a particular important family of weights
was a major breakthrough in the field. Moreover, Novikov’s formula was proven for a somewhat
larger class of weight functions using a completely different and more straightforward method.
In spite of recent progress, various approximate methods have been proposed and are still used
to solve the problem of reconstructing an object from its measured projections.
Some of the first attenuation correction techniques used clinically in nuclear medicine,
especially for SPECT, were designed to correct radionuclide images reconstructed using analytic
techniques such as filtered backprojection. However, as noted above, it is difficult, if not
impossible, to compensate the SPECT image for nonuniform attenuation using analytic
reconstruction techniques. It is possible, however, to correct the radionuclide data assuming that
the attenuation is uniform, for example in the head or pelvis (if the bony structure and other nonuniformities are assumed to be negligible). However, the assumption of a uniformly attenuating
medium is not suitable for reconstructing emission data from the thorax due to the presence of
the lungs. Attenuation correction of radionuclide data using a uniform attenuation map therefore
is not appropriate for myocardial perfusion imaging where photon attenuation is considered to
be a major source of false positive errors.
In cases where the linear attenuation coefficient is uniform throughout the reconstruction
volume, it is possible to reconstruct the radionuclide image with a uniform attenuation map
using an analytic reconstruction technique. This can be accomplished using a method referred to
as the “multiplicative Chang technique” (5). Specifically, if the object has a constant linear
attenuation coefficient within its borders, then the radionuclide image fAC(x,y) corrected for
photon attenuation is given by
yxf
M
i
),(
l
)exp(
1
i
AC
GSPant\Newbook\Final-2008\21-chp\331
1
M
yxf
),(
(3)
Соседние файлы в папке Библиотека им академика М.И. Перельмана