Accurate Monte Carlo Modeling and Performance Evaluation of a Total-body PET Scanner

Abstract

Current clinical PET scanners have poor sensitivity as they detect only ~1% of all emitted coincidence photons owing to the limited axial field-of-view (15 to 25 cm), To overcome this limitation, the EXPLORER consortium has produced the world’s first total-body PET/CT scanner with ~2m field-of-view. In this work, we modeled the geometry of this scanner using the GATE simulation toolkit and then calculated its performance parameters using the NEMA standard. Furthermore, the effect of the time coincidence window and maximum ring difference on count rate and noise equivalent count rate (NECR) were calculated. The results demonstrated that the differences between simulation and experimental sensitivity for the NEMA phantom and extended phantom at the center of FOV were 5 % and 6 %, respectively. They were about 1.5 % for the peak NECR using the extended phantom. The average simulated results of the tangential and radial spatial resolution of point sources were 3.1 mm and 2.74 mm, where the experimental results were 3 mm and 3 mm, respectively. Images of the mini-Derenzo phantom in both simulation and experimental study indicated that rods with a diameter of 2.4 mm could be seen. Using variable time coincidence window, which is a function of the ring difference, improved the NECR by about 10% compared to the constant time coincidence window