摘要
为了提高多能谱X线计算机断层扫描(Computed Tomography,CT)重建算法ASD-NC-POCS的收敛速度,以应用于显微CT双能成像,本文对ASD-NC-POCS算法的迭代过程进行了改进,考虑两个能谱下数据的相关性,建立两个基图像更新方程来更新基图像,并将算法中代数重建方法替换成有序子集联立代数重建方法。为了验证算法可行性,采用两次连续扫描的方式获取双能数据,在实际显微CT系统上进行了小鼠肺部碘造影剂分离的双能成像实验。实验结果表明,相比于ASD-NC-POCS算法迭代300次,改进ASD-NC-POCS方法迭代5次就能从小鼠肺部有效分离出碘造影剂,加快了算法的收敛速度,能有效应用在实际显微CT双能成像中。
New iteration procedures were proposed to improve the convergence speed of multi-spectral computed tomography(CT) reconstruction algorithm of ASD-NC-POCS for dual-energy micro-CT imaging. Specifically, the correlation of data under two energy spectra was considered, two basis images update equations were added to update the basis image, and the ART method in the algorithm was replaced with OS-SART. To verify the feasibility of the proposed method, the dual-energy imaging experiment on the practical micro-CT system was introduced to separate the iodine contrast agent from the lung tissue of model mice. The results showed that the improved ASD-NC-POCS method effectively separates the iodine contrast agent from the lung of model mice with 5 iterations, while its convergence rate was greatly accelerated compared with the original one iterates 300 times. It is demonstrated that the new method can be effectively applied in actual dual-energy micro-CT system for imaging.
作者
雷阳
连茂森
罗守华
LEI Yang;LIAN Maosen;LUO Shouhua(School of Biological Science and Medical Engineering,Southeast University,Nanjing Jiangsu 210096,China)
出处
《中国医疗设备》
2019年第12期54-60,共7页
China Medical Devices
基金
国家重点研发计划（2017YFA0104302）
国家自然科学基金（61871126）
