On-line PET monitoring of radiotherapy beams: image reconstruction and Monte Carlo simulations of detector geometries

The image reconstruction algorithm and design considerations for an annihilation gamma-ray coincidence detection camera are presented. The imaging system is designed to provide fast, on-line monitoring of the positron-emitting activity created by narrow gamma-ray, proton, or heavy-ion radiotherapy beams in tissue for determining total dose, performing range verification for charged particles, and for determining the oxygen content as a function of depth. A model of the detection system was developed and implemented using Monte Carlo techniques. Simulations based on this model are compared with experimental results from a prototype system and are used to study the detection efficiency and imaging characteristics of the system, while parameters such as scintillation crystal spacing, length, and separation are varied. In addition the effects of intercrystal shielding, extended activity distributions, and attenuation on detection efficiency and image reconstruction are presented. Simulation results suggest that the detection geometry which optimizes the sensitivity and spatial resolution utilizes tightly packed, narrow, uncoupled scintillation crystals with no intercrystal shielding. The one-dimensional reconstruction algorithm requires that the angular acceptance be minimized to reduce blurring for extended activity distributions. It is also shown that the depth-of-interaction blurring effect present in annular Positron Emission Tomography imaging systems may be removed through the calibration procedure and the one-dimensional image reconstruction algorithm used for this system