能谱CT通过探测X射线中各能量成分的衰减情况进行断层图像重建，可获得更精准的影像信息，已广泛用于现代医学临床诊断。如何通过低剂量照射获得多种能量分辨区间的能谱探测技术是目前该领域的研究难点。本课题拟研究一种利用侧边界方向入射结构进行分层探测和重构解析X射线能谱的理论和方法。拟从Edge-on型探测器体素单元结构、体素噪声建模分析和混合信号读出降噪技术、可重构能谱解析计算方法等三个方面来展开。课题创新之处在于：1）提出了利用Edge-on型的体素单元结构分层探测并利用电荷耦合方式实现光生电荷的转移读出；2）提出了采用数字域和模拟域混合累加的读出电路进行降噪处理并实现A/D转换；3）根据X射线在半导体材料中的分层吸收理论，重构出不同的能量解析方程组，可以快速求解出多组能量组成成分。课题拟突破传统X射线能谱探测器能量分辨区间固定问题的瓶颈，形成一套新型能量积分式可重构的X射线能谱探测方法学。

Spectral CT, which reconstructs tomographic image by analyzing degradation level of different energy components in x-rays, is able to obtain more precise tissue information and has been widely used in clinical diagnosis. The study of spectral-unmixing-aimed x-ray detector has become a major research focus in medical CT area. This project is going to study an edge-on based reconfigurable voxel structure as well as spectral unmixing methods. The project will be developed in the following aspects, including edge-on voxel structure, voxel noise modeling with its mixed-readout-based denoising technology and reconfigurable energy resolving method. Finally, we will form a novel medical-CT-imaging-aimed energy-integrating spectral unmixing methodology. Our innovation points include (1) proposed an edge-on voxel structure to enable stratified detection and charge-coupling based charge transferring and readout, (2) proposed a mixed AD readout and denoising method and (3) proposed a spectral unmixing method, in which energy-resolving equation-sets are reconfigured based on x-ray absorption effect, to solve multi groups of energy bins.