Validation of a Monte Carlo model of the uEXPLORER total-body PET scanner using GATE code

Abstract

PurposeTo develop and validate a GATE model of the newly manufactured total-body PET scanner (uEXPLORER), in order to evaluate the capability of GATE to simulate PET scanners with long AFOV, and to provide a reliable model to be used in further optimization studies. MethodsThe detector geometry of our model consists of 564,480 LYSO crystals grouped into blocks and modules in accordance with the actual scanner's specifications. Signal processing chain was modeled directly in GATE through the digitizer modules. For model validation, we performed spatial resolution, sensitivity, count rate and scatter fraction tests following the National Electrical Manufactures Association (NEMA) NU 2–2018 protocol, in addition to other devised tests representative of total-body acquisitions. The simulation results were then compared to published measurements. ResultsOverall, our simulation model demonstrated good agreement with the experimental results, within 6% for sensitivity, 7% for scatter fraction and 17% for noise equivalent count rate. For the spatial resolution test, the model matched the experimental results to within 0.1–1.3 mm. ConclusionsWe have shown that our developed model predicts the uEXPLORER PET scanner's performances with good accuracy. Consequently, we conclude that GATE is reasonably suitable for simulating long AFOV PET scanners.