TY - JOUR T1 - Comparative Study of Various Reference Regions for Standardized Uptake Value Ratio Using<sup>11</sup>C-Pittsburgh Compound-B PET and MR Images JF - Journal of Nuclear Medicine JO - J Nucl Med SP - 3099 LP - 3099 VL - 61 IS - supplement 1 AU - Tomohiro Tada AU - Yoshinori Ito AU - Rina Murayama AU - MIka Tamura AU - Yoshinori Tsutsumi AU - Katsuhiko Kato AU - Yuka Ochi AU - Haruna Ikeda Y1 - 2020/05/01 UR - http://jnm.snmjournals.org/content/61/supplement_1/3099.abstract N2 - 3099Objectives: Previously, we compared the suitability of the cerebellum and pons as a reference region for quantitative evaluation on brain PET images in Alzheimer’s disease (AD) using 11C-Pittsburgh Compound-B (11C-PiB). However, we set the volumes of interest (VOIs) manually in the cerebellum and pons as reference regions and frontal and temporal cortex as regions of interest (ROIs). Therefore, it is possible that the VOIs may not have been set accurately. For instance, the regions outside the cortical regions may be contained as specific accumulation areas in VOIs. It is imperative to measure standardized uptake value ratio (SUVR) by setting the VOIs automatically using MRI images and a relevant imaging software. Additionally, SUVR is reportedly better correlated with clinical symptoms when using the white matter as the reference region. Therefore, this study aimed to compare different potential reference regions (the pons, cerebellum, cerebellum + pons or white matter) to identify the most suitable among them using MRI and software. Methods: We retrospectively evaluated the data of 13 AD patients and 17 healthy controls (HCs). We used the PMOD software (PMOD Technologies Ltd.) to set the VOIs in the pons, cerebellum, white matter, frontal cortex, temporal cortex, parietal cortex, and occipital cortex. They were automatically adjusted according to the individual patient’s anatomy, which was visualized via T1-MR images, using a spatial normalization procedure. These VOIs were set to measure the mean of SUV values (SUVmean) and the maximum SUV values (SUVmax) in a VOI. SUVmean and SUVmax values for the pons, cerebellum, cerebellum + pons, and white matter between AD patients and HCs were compared using a t-test. SUVR was then measured using these different reference regions for a variety of ROIs, including the frontal cortex, temporal cortex, parietal cortex, and occipital cortex. These values between AD patients and HCs were subsequently compared using a t-test. Results: Both SUVmean and SUVmax of the pons, cerebellum, cerebellum + pons were similar; however, we noted significant differences between AD patients and HCs in those of the white matter (p&lt;0.01). There were significant differences between the SUVRs of AD patients and HCs in each of the reference regions examined (p&lt;0.01). Therefore, the outcomes suggested that all of these reference regions were suitable to be used for analysis. Furthermore, the greatest difference in values between AD patients and HCs could be identified using SUVmean with the cerebellum as the reference region. Conclusions: The pons, cerebellum, cerebellum + pons, and white matter can all be used as suitable reference regions in the measurement of SUVR. This study also suggested that a more accurate quantitative evaluation can be performed if SUVmean is determined using the cerebellum as the reference region. ER -