像素
探测器
费用分摊
计算机科学
能量(信号处理)
航程(航空)
模拟
发电机(电路理论)
光子计数
电子工程
算法
人工智能
物理
数学
统计
工程类
功率(物理)
电信
航空航天工程
量子力学
作者
David S. Lee,X. Zhan,W. Y. Tai,Wojciech Zbijewski,Katsuyuki Taguchi
标识
DOI:10.1109/nssmicrtsd49126.2023.10337981
摘要
Photon-counting detector (PCD)-based computed tomography (CT) imaging improves diagnostic accuracy and lowers radiation exposure, requiring realistic and versatile simulation tools for algorithm development. Existing model-based PCD simulators, however, have limitations in accurately modeling PCD outputs, such as missing pixel-to-pixel variations and unidentified mismatches between ideal models and actual measurements. The study aimed to develop an advanced PCD-based CT simulator that overcomes these limitations in existing simulators. Our simulator is based on the PCD output model that was built upon the cascaded model. To develop an advanced PCD simulator, we modified and generalized the previously established condition-specific functions, creating a count-rate-dependent energy mapping operator and a pixel-to-pixel variation generator. These components were then incorporated into the existing model. To evaluate the simulator's performance, we compared its outputs with CdZnTe-based PCD measurements and analyzed the effects of varying several tunable parameters on spectrum outputs and pixel-to-pixel variations. Comparison of the simulator outputs with the measurements showed remarkable agreement for a wide range of incident count-rates. The mean absolute percentage error for all energy bins was 4.67 %, indicating effective reproduction of the physical PCD measurements. Adjusting the major parameters visually showed their spectral distortion effects due to pulse pileup and charge sharing. This adjustment also allowed us to easily modulate or balance the pixel-to-pixel variations across energy bins, indicating effective control of the output spectrum and these variations. Our study successfully developed an advanced PCD simulator that offers a more accurate representation of the inherent detector behavior. The proposed simulator is a valuable tool for developing and evaluating new algorithms and techniques for PCD-based CT imaging.
科研通智能强力驱动
Strongly Powered by AbleSci AI