Iterative feature refinement for accurate undersampled MR image reconstruction
Wang, Shanshan; Liu, Jianbo; Liu, Qiegen; Ying, Leslie; Liu, Xin; Zheng, Hairong; Liang, Dong
刊名PHYSICS IN MEDICINE AND BIOLOGY
2016
英文摘要Accelerating MR scan is of great significance for clinical, research and advanced applications, and one main effort to achieve this is the utilization of compressed sensing (CS) theory. Nevertheless, the existing CSMRI approaches still have limitations such as fine structure loss or high computational complexity. This paper proposes a novel iterative feature refinement (IFR) module for accurate MR image reconstruction fromundersampled K-space data. Integrating IFR with CSMRI which is equipped with fixed transforms, we develop an IFR-CS method to restore meaningful structures and details that are originally discarded without introducing too much additional complexity. Specifically, the proposed IFR-CS is realized with three iterativesteps, namely sparsity-promoting denoising, feature refinement and Tikhonov regularization. Experimental results on both simulated and in vivo MR datasets have shown that the proposed module has a strong capability to capture image details, and that IFR-CS is comparable and even superior to other state-of-the-art reconstruction approaches.
收录类别SCI
原文出处http://iopscience.iop.org/article/10.1088/0031-9155/61/9/3291/meta
语种英语
内容类型期刊论文
源URL[http://ir.siat.ac.cn:8080/handle/172644/10454]  
专题深圳先进技术研究院_医工所
作者单位PHYSICS IN MEDICINE AND BIOLOGY
推荐引用方式
GB/T 7714
Wang, Shanshan,Liu, Jianbo,Liu, Qiegen,et al. Iterative feature refinement for accurate undersampled MR image reconstruction[J]. PHYSICS IN MEDICINE AND BIOLOGY,2016.
APA Wang, Shanshan.,Liu, Jianbo.,Liu, Qiegen.,Ying, Leslie.,Liu, Xin.,...&Liang, Dong.(2016).Iterative feature refinement for accurate undersampled MR image reconstruction.PHYSICS IN MEDICINE AND BIOLOGY.
MLA Wang, Shanshan,et al."Iterative feature refinement for accurate undersampled MR image reconstruction".PHYSICS IN MEDICINE AND BIOLOGY (2016).
个性服务
查看访问统计
相关权益政策
暂无数据
收藏/分享
所有评论 (0)
暂无评论
 

除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。


©版权所有 ©2017 CSpace - Powered by CSpace