基于残差图分解与平滑块排列的低剂量CT投影降噪算法  

Sinogram Noise Reduction for Low-dose CT Based on Residual Image Decomposition and Smooth Patch Ordering

作　　者：王燕玲[1,2] 薛敏 邵燕灵 刘祎[1,2] 张权[1,2] 陈燕[1,2] 桂志国[1,2] WANG Yan-ling;XUE Min;SHAO Yan-ling;LIU Yi;ZHANG Quan;CHEN Yan;GUI Zhi-guo(Key Laboratory for Biomedical Imaging and Big Data of Shanxi Province, North University of China, Taiyuan 030051, China;School of Information and Communication Engineering, North University of China, Taiyuan 030051, China;School of Science, North University of China, Taiyuan 030051, China)

机构地区：[1]中北大学生物医学成像与影像大数据山西省重点实验室,山西太原030051 [2]中北大学信息与通信工程学院,山西太原030051 [3]中北大学理学院,山西太原030051

出　　处：《中北大学学报（自然科学版）》2018年第2期170-176,共7页Journal of North University of China(Natural Science Edition)

基　　金：国家重大科学仪器设备开发专项资助项目(2014YQ24044508);国家自然科学基金资助项目(61671413;61771433);电子测试技术重点实验室开放基金资助项目(ZDSYSJ2015006);山西省回国留学人员科研资助项目(2016-085);山西省自然科学基金资助项目(201601D011003);中北大学校基金资助项目(XJJ2016019)

摘　　要：针对低剂量CT(Computed Tomography)图像质量严重退化的问题,将平滑块排列(Smooth Patch Ordering,SPO)算法应用于低剂量CT投影降噪中,并通过残差图分解提取图像结构信息以进一步提高SPO算法的去噪性能.首先对低剂量CT投影进行SPO滤波得到初始的降噪投影,然后通过改进的形态分量分析(Morphological Component Analysis,MCA)方法将残差图分解为结构部分和噪声部分,并将残差图中提取的结构部分作为补偿投影,与降噪投影相加得到新的降噪投影.最后利用滤波反投影(Filtered-Back Projection,FBP)方法对新的降噪投影进行重建,获得重建的低剂量CT图像.实验结果证明,与其他传统算法相比,所提算法在低剂量CT重建图像的噪声抑制以及边缘、细节和结构等特征的保持方面都有明显的优越性.Aiming at the problem that low-dose CT（Computed Tomography）images tend to be severely degraded,an algorithm of smooth patch ordering（SPO）is applied to sinogram noise reduction for lowdose CT.Additionally,the performance of the SPO algorithm is further improved by decomposing the residual image and extracting the structure information of the image.Firstly,the SPO approach is applied to sinogram filtering for low-dose CT to produce an initial denoised projection.Then,the residual image is decomposed into the structure and the noise parts by the improved morphological component analysis（MCA）method.The structure part extracted from the residual image,as a compensation image,is added back to the denoised projection to generate a new denoised projection.Finally,the reconstructed image for low-dose CT is obtained by performing filtered-back projection（FBP）from the new denoised projection.The results from computer experiments demonstrate that the proposed method performs better not only in noise suppression,but also in edge,detail and structure preservation,when compare with traditional reconstruction algorithms for low-dose CT.

关 键 词：低剂量CT 形态分量分析 平滑块排列 残差图分解 投影降噪 结构保持 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]