Development of a Monte Carlo simulation platform for the systematic evaluation of photon-counting detector-based micro-CT
Shiyan Yang,Mengjia Xue,Tianwu Xie
DOI: https://doi.org/10.1016/j.ejmp.2024.104824
Abstract:Purpose: This study aimed to develop a photon-counting detector (PCD) based micro-CT simulation platform for assessing the performance of three different PCD sensor materials: cadmium telluride (CdTe), gallium arsenide (GaAs), and silicon (Si). The evaluation encompasses the components of primary and scatter signals, performance of imaging contrast agents, and detector efficiency. Methods: Simulations were performed using the Geant4 Monte Carlo toolkit, and a micro-PCD-CT system was meticulously modeled based on realistic geometric parameters. Results: The simulation can obtain HU values consistent with measured results for iodine and calcium hydroxyapatite contrast agents. The two major components of scatter signals for CdTe and GaAs based PCD are fluorescent X-ray photons and photoelectrons, whereas for Si, the components are photoelectrons and Compton electrons. Scattering counts of CdTe and GaAs sensors can be effectively reduced by using energy thresholds, whereas those of Si sensor are insensitive to the applied threshold. The optimal threshold values for CdTe and GaAs are 30 and 15 keV, respectively. For contrast agent imaging, GaAs exhibits enhanced sensitivity to low photon energies compared to CdTe, while it's contrast-to-noise ratio (CNR) values are slightly lower than those of CdTe at the same contrast agent concentration. Among the three sensor materials, Si has the lowest CNR and detector efficiency; CdTe exhibits the highest efficiency, except in low-energy ranges (< 45 keV), where GaAs has superior efficiency. Conclusions: The proposed methods are expected to benefit PCD optimization and applications, including energy threshold selection, scattering correction, and may reduce the need for large-scale experiments.