Coincidence time resolution measurements for dual-ended readout PET detectors

Abstract

Coincidence time resolution (CTR) is an important performance parameter of clinical positron emission tomography (PET) detectors, which is dependent on the parameters of the scintillation crystal, the timing measurement methods, and the electronics. In this work, CTRs of PET detectors using dual-ended readout of segmented LYSO arrays were measured with different timing measurement methods. The detector modules were composed of LYSO arrays of 8 × 8 with a crystal size of 3.10 × 3.10 × 20 mm3 or 13 × 13 with a crystals size of 1.88 × 1.88 × 20 mm3, coupled at each end to an 8 × 8 Hamamatsu SiPM array (S13361-3050NE-08) with pixel active area of 3.00 × 3.00 mm2 and pitch of 3.20 mm. The signals of the SiPM arrays were individually read out and processed by using the TOFPET2 ASICs. The CTRs of the detectors were measured after the timing alignment of the ASICs was performed. Timing measurement methods of the dual-ended readout detector such as the average time of the two SiPMs, the faster time of the two SiPMs, the time of the back SiPM, and the time of the front SiPM were compared. The average time of the two SiPMs provides the best CTR, which is explained by a simple model about the timing measurement methods of a dual-ended readout detector. The detector with one-to-one crystal-to-SiPM coupling provides better CTR than the detector with a crystal size smaller than the pixel size of the SiPM. The best CTR of 367 ± 6 ps was obtained by using the 8 × 8 LYSO array with unpolished lateral crystal surfaces and ESR reflectors. The CTR of the dual-ended readout detector can be improved if a DOI-dependent timing correction is used.