Page 366 - Biomedical Engineering and Design Handbook Volume 2, Applications
P. 366

344  DIAGNOSTIC EQUIPMENT DESIGN

                         The reconstructed spatial resolution is approximately 8 to 15 mm in a PET or SPECT clinical
                       study. To achieve similar sampling of the anatomy of the rat and mouse, the spatial resolution would
                       have to be 1.7 and 0.75 mm, respectively. Unfortunately, spatial resolution is not the sole factor in
                       determining image quality. Adequate count density levels have to be achieved so that the statistical
                       fluctuations do not overwhelm the information content. Count density comparisons of the associated
                       resolution volumes (0.75 mm vs. 8 mm) suggests that the sensitivity of a small animal scanner needs
                       to be more than 2000 times higher than of a human scanner. This is currently far out of reach, but
                       the small size of the animals has one big advantage. In humans, attenuation reduces the coincidence
                       event rate by more than a factor of 50 and in SPECT image by at least a factor of 10. In the rat and
                       mouse, the attenuation reduction factors are less than 2. When attenuation is also considered, the
                       required sensitivity increase is approximately 5 for rat imaging and 55 for mouse imaging. Another
                       gain comes improving the geometric efficiency of the detecting systems by making them just large
                       enough to accommodate rats. Because other compromises required to improve spatial resolution are
                       employed such as small pinholes for SPECT or thinner detectors for PET, the overall sensitivity gain
                       is less than a factor of 5. Thus, we approach the sampling and signal-to-noise limits in animals
                       smaller than rats. To some degree, further gains can be realized by increasing the amount of activity
                       administered to the animals. That also has constraints associated with the count rate capability of the
                       scanner and the volume of injectate tolerated by the animal.
                         Concerns about the limited anatomical information associated with SPECT and PET are also
                       associated with small animal imaging. As a result, most SPECT and PET small animal imaging sys-
                       tems also include high-resolution CT imaging systems, with some of them capable of spatial reso-
                       lution better than 10 μm.


           11.5.1 Small Animal PET
                       Many investigators have described small animal PET systems and there are currently more than four
                       commercial systems available. By and large, small animal PET systems are scaled-down versions of
                       clinical PET scanners. These are based on multiple rings of discrete scintillators with LSO, LYSO,
                       or GSO being the most commonly used detectors. To achieve good spatial resolution, the face size
                       of the crystal is small, ranging from less than 1 × 1 mm in some experimental systems up to 2 × 2 mm.
                       The detectors are 10 to 15 mm thick to limit depth of interaction losses. The spatial resolution in
                       these systems ranges from 1.2-to 2.2-mm FWHM. The ring diameter is 15 to 20 cm and the overall
                       detection efficiency is on the order of 2 to 5 percent. Typical small animal PET images are shown in
                       Fig. 11.19a.



           11.5.2 Small Animal SPECT
                       Because the spatial resolution of SPECT systems designed for human imaging is fairly coarse, it would
                       seem unlikely that SPECT could ever serve for imaging small animals. What makes small animal
                       SPECT possible is pinhole collimation. Although pinhole imaging has low-count sensitivity for imaging
                       large distributed sources, pinholes become advantageous when imaging localized distributions. Through
                       magnification, pinhole imaging reduces the apparent intrinsic spatial resolution of the detector, result-
                       ing in an overall spatial resolution that is predominantly determined by the diameter of the pinhole
                       aperture. Spatial resolution below 2-mm FWHM is routinely achieved with both investigational and
                       commercial devices (see example images in Fig. 11.20), and submillimeter resolution can be achieved
                       by concentrating on small-source distributions within the animal. The accurate modeling of the pin-
                       hole aperture is a crucial factor in both the design of the collimators and the reconstruction algorithms
                       and has lead to a number of papers reanalyzing the physics of pinhole imaging. 42–45
                         Although spectacular spatial resolution can be achieved through magnification with pinhole
                       collimation, it comes with the very real price of either very large detectors or limited field of view.
                       The need for magnification is reduced for detectors with good intrinsic spatial resolution. While most
                       of the currently available detectors have intrinsic spatial resolution in the range of 2.5 to 3.5 mm, it
   361   362   363   364   365   366   367   368   369   370   371