Characterization of Depth of Interaction Estimation Accuracy for 3-layer Depth of Interaction Positron Emission Tomography Detector

This study proposed a three-layer depth of interaction (DOI) detector utilizing a quasi-monolithic crystal to address spatial resolution degradation caused by parallax errors in positron emission tomography (PET) detectors. Monte Carlo simulations were performed to analyze optical signals, and the modified resistive charge division method, combined with the maximum likelihood position estimation (MLPE) algorithm, was used for analysis and evaluation. The 2D flood histogram results indicated that light spread in the quasi-monolithic layer with layer thickness, complicating DOI identification. The MLPE results demonstrated that as the thickness of the quasi-monolithic layer increases, DOI identification accuracy exceeding 70% was achieved, with the highest accuracy of 84.2% observed at a total thickness of 7 mm. These findings show that the three-layer DOI-PET detector using a quasi-monolithic crystal is capable of accurate three-layer DOI identification.