Design and performance evaluation of spheroid geometry for brain PET scanner using Monte Carlo modeling

Introduction: There has been a curiosity about the spheroid geometry for PET scanners developments since several years ago, therefore in this study, we are aiming to evaluate the performance of this geometry and compare its performance with cylindrical geometry using Monte Carlo simulation. Methods: We simulated a spheroid geometry with a radius of 199 mm, patient bore with of radius of 175 mm, which is compatible with brain size. In second design, cylindrical geometry was simulated with transaxial FOV and ring radius of 175 mm as well. Photon detection efficiency (PDE), NEMA line source sensitivity, spatial resolution and Derenzo phantom image quality were analyzed. Results: We obtained PDE about 21.7% versus 23.8% in 250-750 keV and 19.5% versus 21.3% in 410-613 keV for point source in center of FOV for spheroid and cylindrical PET respectively. The results of NEMA sensitivity measurements indicate 3.29 kcps/MBq versus 3.64 kcps/MBq for spheroid and cylindrical designs. The spatial resolution (FWHM) calculations using MLEM reconstruction algorithm show around 1.6 mm for transvers and axial resolution for point source placed in center of FOV for both scanners. Also we found for spheroid and cylindrical designs 4.8 and 2.7 mm versus 4 and 3.6 mm as transvers and axial mean resolution for off-center point sources. Conclusion: Performance evaluation study indicates that the spheroid geometry delivers better axial resolution whereas cylindrical design can still provide higher sensitivity and transvers spatial resolution than the spheroid geometry PET with same scanner bore size.