A Detector Design For Pet Using Monte Carlo Simulation Based On Sipm And Continuous Crystal Of Lyso

Positron Emission Tomography (PET) is a kind of non-invasive imaging technology, which can provide three-dimensional image of radioactive tracer distribution. Detector design is one of the most important parts in PET development.A detector module has two main components which are scintillator crystal and photon detector. The novel photon detector, SiPM is getting increasing attention because it is compact and provides high gain. Continuous scintillator crystal can decrease the cost and has higher sensitivity comparing with discrete scintillator crystal. Thus, using continuous scintillator crystal is the trend of PET detector design. LYSO has fine physics and optical performance. Therefore, it is possible to get a PET detector with high sensitivity, considerable spatial resolution and low cost using SiPM and continuous scintillator crystal of LYSO.This thesis aims at designing a PET detector with intrinsic spatial resolution of 2mm. It has discussed the detector design using SiPM and continuous scintillator crystal of LYSO. At the beginning, a few tentative designs were set and were simulated roughly. As a result, we could get several kinds of designs which were in line with the request. Then detailed simulation was applied to evaluate the performance of designs we got. At the fist step, crystal with three kinds of thickness were simulated using DETECT2000 (a program for simulating) - 5mm, 10mm, 15mm, and there were four kinds of surface treatment: METAL 0.20RC, METAL 0.98RC, POLISH 0.20RC, POLISH 0.98RC. For the thickness of 5mm, the position sensitivity for four kinds of treatment were too bad as photons hadn't enough diffusion resulting from the small thickness of crystal, and the spatial resolution fluctuated greatly too. For the thickness of 10mm, the position sensitivity for four kinds of treatment was fine and the spatial resolution was less than 1mm. Specially, the treatment of METAL 0.20RC got the best spatial resolution. For the thickness of 15mm, the position sensitivity for four kinds of treatment was considerable, but the spatial resolution was too bad. At the second step, crystal with two kinds of thickness: 5mm&10mm and one kind of surface treatment METAL 0.20RC were simulated using both GATE (a program for simulating) and DETECT2000. Comparing results of the two settings, we found the setting with thickness of 5mm got better performance. This setting had considerable position sensitivity and the spatial resolution was less than 1.2mm.