TY - JOUR T1 - Optimization of Standardized Uptake Value of Fluorodeoxyglucose for Cortical Extraction in dedicated brain PET System JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 3024 LP - 3024 VL - 62 IS - supplement 1 AU - Daisuke Morimoto-Ishikawa AU - Kohei Hanaoka AU - Shota Watanabe AU - Atsushi Otani AU - Tetsuro Mizuta AU - Yoshiyuki Yamakawa AU - Hayato Kaida AU - Kazunari Ishii Y1 - 2021/05/01 UR - http://jnm.snmjournals.org/content/62/supplement_1/3024.abstract N2 - 3024Introduction: In the diagnosis of Alzheimer’s disease and other neurodegenerative diseases, detecting the typical regional patterns of glucose metabolism abnormalities by 118F-fluorodeoxyglucose (FDG) PET brain imaging is playing an important role. The spatial resolution of PET is intrinsically limited by some physical factors, which may lead to substantial underestimation or overestimation. The purpose of this study is to compare the optimal threshold standardized uptake value (SUV) for cerebral cortical extraction of images obtained by a dedicated brain PET system with those by a whole-body PET system. Methods: Four patients who underwent both brain FDG PET and 3D T1-weighted MRI for the diagnosis of dementia were included. The PET images were acquired continuously for 30 minutes each using whole-body PET system (Discovery 710, GE Healthcare) and newly developed dedicated brain PET system (SET-5002, Shimadzu Corporation) at 60 and 90 minutes after FDG administration. The full width at half maximum at the center of the detector for each system were 5.27 and 2.45 mm, respectively. The workstation (Advantage Workstation 4.7, GE Healthcare) was used for semi-automated segmentation technique to measure the maximum SUV (SUVmax) of the cerebral cortex at the level of basal ganglia and the mean SUV (SUVmean) of the cerebellum. Appropriate threshold value was determined so that the volume of PET delineation fits to an actual cerebral cortex volume on MRI as the reference standard, and the percentage errors of the threshold compared to the measured SUVmax or SUVmean were calculated. Results: Both SUVmax andSUVmean obtained by the dedicated brain PET system tended to be higher than those obtained by the whole-body PET system (median SUVmax [range]: whole-body PET system, 11.86 [10.67-12.59]; dedicated brain PET system, 18.65 [17.50-21.30]. median SUVmean[range]: whole-body PET system, 5.77 [4.94-5.92]; dedicated brain PET system 8.45 [7.75-8.75]). As for the threshold for cortical extraction, the percentage error for dedicated brain PET system tended to be lower than that for whole-body PET system when the SUVmax was used (median percentage error[range]: whole-body PET system, 53.6% [48.0-56.6%]; dedicated brain PET 49.5% [44.0-51.8%]), and the threshold for dedicated brain PET system tended to be higher than that for whole-body PET when SUVmean was used (median percentage error[range]: whole-body PET system, 106.0% [100.0-122.2%]; dedicated brain PET 112.1% [107.7-123.6%]). Conclusions: For cortical extraction using dedicated brain PET system, which has high-resolution, a slightly higher rate is required for the threshold based on SUVmax in the cerebral cortex, and a lower rate is required for based on SUVmean in the cerebellum. ER -