TY - JOUR T1 - Performance Assessment of a Preclinical PET Scanner with Pinhole Collimation by Comparison to a Coincidence-Based Small-Animal PET Scanner JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 1368 LP - 1374 DO - 10.2967/jnumed.113.136663 VL - 55 IS - 8 AU - Matthew D. Walker AU - Marlies C. Goorden AU - Katherine Dinelle AU - Ruud M. Ramakers AU - Stephan Blinder AU - Maryam Shirmohammad AU - Frans van der Have AU - Freek J. Beekman AU - Vesna Sossi Y1 - 2014/08/01 UR - http://jnm.snmjournals.org/content/55/8/1368.abstract N2 - PET imaging of rodents is increasingly used in preclinical research, but its utility is limited by spatial resolution and signal-to-noise ratio of the images. A recently developed preclinical PET system uses a clustered-pinhole collimator, enabling high-resolution, simultaneous imaging of PET and SPECT tracers. Pinhole collimation strongly departs from traditional electronic collimation achieved via coincidence detection in PET. We investigated the potential of such a design by direct comparison to a traditional PET scanner. Methods: Two small-animal PET scanners, 1 with electronic collimation and 1 with physical collimation using clustered pinholes, were used to acquire data from Jaszczak (hot rod) and uniform phantoms. Mouse brain imaging using 18F-FDG PET was performed on each system and compared with quantitative ex vivo autoradiography as a gold standard. Bone imaging using 18F-NaF allowed comparison of imaging in the mouse body. Images were visually and quantitatively compared using measures of contrast and noise. Results: Pinhole PET resolved the smallest rods (diameter, 0.85 mm) in the Jaszczak phantom, whereas the coincidence system resolved 1.1-mm-diameter rods. Contrast-to-noise ratios were better for pinhole PET when imaging small rods (<1.1 mm) for a wide range of activity levels, but this reversed for larger rods. Image uniformity on the coincidence system (<3%) was superior to that on the pinhole system (5%). The high 18F-FDG uptake in the striatum of the mouse brain was fully resolved using the pinhole system, with contrast to nearby regions equaling that from autoradiography; a lower contrast was found using the coincidence PET system. For short-duration images (low-count), the coincidence system was superior. Conclusion: In the cases for which small regions need to be resolved in scans with reasonably high activity or reasonably long scan times, a first-generation clustered-pinhole system can provide image quality in terms of resolution, contrast, and the contrast-to-noise ratio superior to a traditional PET system. ER -