基于联合正则化及压缩传感的MRI图像重构.pdfVIP

到鸟哮 章良功、 第36卷第3期 Vo l. 36 No.3 2010 年 5 月 OPTICAL TECHNIQUE 岛1ay 2010 文章编号: 1002-1582(2010)03-0350-06 基于联合正则化及压缩传感的 MRI 图像重构提 王艳，练秋生，李凯 (燕山大学信息科学与工程学院，向北秦皇岛 066004) 摘 要:基于压缩传感的 MRI 图像重构利用图像稀疏的先验知识能从很少的投影值重构原图像。目前 MRI 重构 算法只利用 MRI 图像稀疏性表示或只利用基于其局部光滑性的先验知识，重构效果不理想。针对此问题，结合两种先 验知识，提出一种基于联合正则化及压缩传感的 MRI 图像重构方法。利用块坐标下降法将求解联合正则化问题转化为 交替求解二次凸优化、稀疏正则化和全变差正则化三个简单的优化问题。并提出分别采用共辄梯度法、二元自适应收缩 法以及梯度下降法对以上优化问题求解。实验结果表明，该算法重构效果比现有算法有明显地提高。 关键词: MRI 图像重构;压缩传感;联合正则化;共辄梯度法 中图分类号: 1、P391 文献标识码:A MRI reconstruction based on compound regularizers and compressed sensing WANG Yan , LlAN Qiu….sheng , LI Kai CInstitute of Information Science and Technology , Yanshan University , Qinhuangdao 066004 , China) Abstract: The reconstruction of MRI based on compressed sensing is able to recover the original image from the fe• wer projections using the sparse priors of image. The current MRI reconstruction algorithms simply use either the sparse priors or the local smooth priors of MRI image , and result in the inferior reconstructio n. Pointing at this problem , the MRI reconstruction algorithm based on the compound regularizers and compressed sensing which uses the two priors sim俨 ultaneously is proposed. Applyìng the block-coordìnate descent method , the solutìon to compound regularizers is turned into dealing with the following three simple optimization problems alternatively: the quadratic convex optimization , the sparsity regularizer and the total vari