平滑硬阈值约束法重建有限角度CT图像  

Smooth Hard Threshold Regularizer to Reconstruct Limited-Angle CT Image

作　　者：陈梓嘉 黄少福[1] 许少生[1] CHEN Zi-jia;HUANG Shao-fu;XU Shao-sheng(Department of Radiation Oncology,Shantou Central Hospital,Shantou 515031,China)

机构地区：[1]汕头市中心医院肿瘤放疗科,汕头515031

出　　处：《核电子学与探测技术》2023年第2期279-282,共4页Nuclear Electronics & Detection Technology

基　　金：汕头市医疗卫生科技计划(220428156491494)资助。

摘　　要：针对有限角度CT图像重建问题,探讨了一种基于图像对称的平滑硬阈值约束的CT重建算法。通过图像的局部轴对称变换,利用非伪影区域来恢复重建图像的伪影区域,然后采用平滑硬阈值方法校正非伪影区域带来的伪结构信息。新算法主要包含四个步骤:代数重建、非负修正、局部轴对称变换及SHT正则化,循环执行上述步骤直至满足迭代收敛条件。重建结果表明,该算法可有效消除伪影并还原图像细节,显著改善图像质量。To solve the problem of limited-angle CT image reconstruction,a CT reconstruction algorithm based on image symmetry of smooth hard threshold(SHT)regularization is discussed.In this method,a local axisymmetric transformation of image is designed to restore the pseudo-shadow area of reconstruction image by the information of non-pseudo shadow area.Moreover,the smooth hard threshold method is used to calibrate the pseudo-structural information of the pseudo shadow area.The proposed algorithm mainly consists of four steps:Algebraic reconstruction technique(ART);Positivity constraint;Local axisymmetric transformation;SHT regularization.The above steps alternated each other until the convergence criterion is satisfied.The simulated reconstruction result of Shepp-Logan phantom demonstrates that the proposed algorithm could eliminate pseudo-shadow of reconstruction image more effectively,restore image details better,and improve the image quality significantly.

关 键 词：计算机断层扫描 有限角度 迭代重建 图像对称 平滑硬阈值 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]