I. Introduction
Alarger number of decodable scintillation crystals per photomultiplier tube (PMT) can increase spatial resolution in PET detectors without an accompanying increase in cost. High resolution crystal identification can be achieved by properly controlling the distribution of light to the PMTs. Many documents about high-resolution crystal decoding have been published in the past decade [1]–[4]. Some crystal decoding even involves multi-layer blocks using position-sensitive PMTs [5], [6].