Inclusion of a Charge Sharing Correction Algorithm Into an X-Ray Photon Counting Spectral Detector Simulation Framework

Most photon counting spectral imaging (x-CSI) simulations combine Monte Carlo techniques and experimental measurements to model signal output. These often omit some physical processes to minimize the computational resources needed, resulting in higher simulated energy resolutions compared with experiments. The aggregate effect of unsimulated processes can be compensated for using experimentally derived “blurring” factors. Blurring factors usually take the form of Gaussian convolutions of the simulation output to model system noise and collection inefficiencies. This approach is not suitable for applications involving charge sharing correction algorithm (CSCA) modeling, however, as most CSCAs operate in the prethresholding domain and across multiple pixels, incorporating variable amounts of noise and collection efficiency. To better model x-CSI CSCAs, this work introduces CoGI: a framework combining Monte Carlo and finite element methods to simulate photon-matter interaction, intrapixel charge transport, interpixel charge sharing, signal generation, and CSCA. CoGI is the first full x-CSI simulation framework to incorporate CSCA reported in the literature. CoGI has been validated in spectral and energy binning modes, using a γ-ray spectrometer and an Acteon series pixelated x-CSI detector, respectively. CoGI thus represents a versatile framework for comparing multiple ASIC-based CSCAs and will help inform future x-CSI system development.