Design and simulation of a novel method for determining depth-of-interaction in a PET scintillation crystal array using a single-ended readout by a multi-anode PMT

Abstract

PET detectors with depth-of-interaction (DOI) encoding capability allow high spatial resolution and high sensitivity to be achieved simultaneously. To obtain DOI information from a mono-layer array of scintillation crystals using a single-ended readout, the authors devised a method based on light spreading within a crystal array and performed Monte Carlo simulations with individual scintillation photon tracking to prove the concept. A scintillation crystal array model was constructed using a grid method. Conventional grids are constructed using comb-shaped reflector strips with rectangular teeth to isolate scintillation crystals optically. However, the authors propose the use of triangularly shaped teeth, such that scintillation photons spread only in the x-direction in the upper halves of crystals and in the y-direction in lower halves. DOI positions can be estimated by considering the extent of two-dimensional light dispersion, which can be determined from the multiple anode outputs of a position-sensitive PMT placed under the crystal array. In the main simulation, a crystal block consisting of a 29 × 29 array of 1.5 mm × 1.5 mm × 20 mm crystals and a multi-anode PMT with 16 × 16 pixels were used. The effects of crystal size and non-uniform PMT output gain were also explored by simulation. The DOI resolution estimated for 1.5 × 1.5 × 20 mm3 crystals was 2.16 mm on average. Although the flood map was depth dependent, each crystal was well identified at all depths when a corner of the crystal array was irradiated with 511 keV gamma rays (peak-to-valley ratio ∼9:1). DOI resolution was better than 3 mm up to a crystal length of 28 mm with a 1.5 × 1.5 mm2 or 2.0 × 2.0 mm2 crystal surface area. The devised light-sharing method allowed excellent DOI resolutions to be obtained without the use of dual-ended readout or multiple crystal arrays.