基于投影域分解的多能谱CT造影剂物质识别研究

Multi-Energy Spectral CT Contrast Agent Material Recognition Based on Projection Domain Decomposition

本文结合造影剂物质的K-edge信息及能谱CT双效应分解模型计算投影域分解系数,应用最小二乘法对多个能谱通道下的投影域进行分解,并重建生成光电效应、康普顿效应及各个造影剂物质所对应的图像,实现造影剂物质的识别.仿真实验表明,对分解后的重建图像衰减值进行分析,能够显示与解剖学背景分离的各个造影剂物质的图像及其浓度,实现造影剂物质的分离成像及对造影剂浓度的定量分析.

The K-edge information of the contrast agent and the dual-effect decomposition model of the energy spectrum CT were used to calculate the decomposition factor of the projection domain,the projection domain under multiple spectral channels was decomposed by the least squares method,and the images of photoelectric effect,compton effect and each contrast agent substance were reconstructed.The identification of the contrast agent substance was achieved.Simulation experiments show that the analysis of attenuation value of the reconstructed image after decomposition can display the image and concentration of each contrast agent separated from the anatomical background,and realize the separation imaging and quantitative analysis of the contrast agent concentration.