基于投影加权的多能光子计数X线CT全能谱图像重建改进方法  

Improved projection-based weighting method for full spectral image reconstruction of multi-energy photon counting X-ray CT

作　　者：周正东[1] 管绍林 余子丽 张雯雯[1] 

机构地区：[1]南京航空航天大学核科学与工程系,南京210016

出　　处：《东南大学学报（自然科学版）》2016年第6期1126-1131,共6页Journal of Southeast University：Natural Science Edition

基　　金：国家自然科学基金资助项目(51575256);中央高校基本科研业务费专项资金资助项目(NP2015101);江苏高校优势学科建设工程资助项目

摘　　要：为了提高多能光子计数X线CT全能谱重建图像的对比噪声比,提出了一种基于投影加权的图像重建改进方法.首先,针对包含不同对比材料的3种模体进行投影仿真,获得包含泊松噪声和高斯噪声的投影正弦图;然后,从投影正弦图中提取噪声信息,构造关于权重的对比噪声比优化函数;最后,求解出使对比噪声比最优的权重,利用该权重进行投影加权图像重建,并对重建图像进行评估.实验结果表明:与当前仅考虑泊松噪声的基于投影加权的图像重建方法相比,对于含钙材料的模体,利用改进方法可明显提高重建图像的对比噪声比;对于含碘材料和等效软组织材料的模体,利用改进方法重建的图像对比噪声比则无显著统计差异.To improve the contrast-to-noise ratio （CNR） of the full spectral reconstructed image of multi-energy photon counting X-ray computed tomography （CT） , an improved projection-based weighting method for image reconstruction is proposed. First, the projection simulation for three kinds of phantom including different contrast materials was carried out and the projection sinograms with Poisson noise and Gaussian noise were obtained. Then, the noise information was extracted from these sinograms, and the CNR optimization function with regard to the weights was formulated. Finally, the optimal weights were figured out and used for the projection-based weighting image reconstruction, and the reconstructed images are evaluated. The experimental results show that, compared with the current projection-based weighting image reconstruction method considering Poisson noise only, the proposed method can obviously improve the CNR of the reconstructed image for the phantom containing contrast material calcium. However, there is no statistically significant difference in CNR for the phantoms containing either iodine or contrast material soft tissue.

关 键 词：光子计数 X线CT 投影 加权 图像重建 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] R318[自动化与计算机技术—计算机科学与技术]