Journal
MEDICAL IMAGE ANALYSIS
Volume 73, Issue -, Pages -Publisher
ELSEVIER
DOI: 10.1016/j.media.2021.102190
Keywords
Compressed sensing; Dynamic MR imaging; Deep learning; Image reconstruction
Categories
Funding
- National Key Research and Development Program of China [2017YFC0108802, 2020YFA0712200]
- National Natu-ral Science Foundation of China [12026603, U1805261, 61771463, 81971611, 61871373, 81729003]
- Key Field R&D Program of Guangdong Province [2018B030335001]
- Science and Technology Plan Program of Guangzhou [2020 07030 0 02]
- Shenzhen Peacock Plan Team Program [KQTD20180413181834876]
- Innovation and Technology Commission of the Government of Hong Kong SAR [MRP/001/18X]
- Engineering Laboratory Program of Chinese Academy of Sciences [KFJ-PTXM-012]
- Strategic Priority Re-search Program of Chinese Academy of Sciences [XDC07040000, XDB25000000]
- China Postdoctoral Science Foundation [2020M682990, 2021M69331]
Ask authors/readers for more resources
In this study, a model-based low-rank plus sparse network (L+S-Net) is proposed for dynamic MR image reconstruction, achieving clear separation of the L component and S component through learned soft singular value thresholding and demonstrating global convergence. Experimental results show that the proposed model outperforms the current state-of-the-art methods on retrospective and prospective cardiac cine datasets, with great potential for high acceleration factors.
In dynamic magnetic resonance (MR) imaging, low-rank plus sparse (L+S) decomposition, or robust principal component analysis (PCA), has achieved stunning performance. However, the selection of the parameters of L+S is empirical, and the acceleration rate is limited, which are common failings of iterative compressed sensing MR imaging (CS-MRI) reconstruction methods. Many deep learning approaches have been proposed to address these issues, but few of them use a low-rank prior. In this paper, a model-based low-rank plus sparse network, dubbed L+S-Net, is proposed for dynamic MR reconstruction. In particular, we use an alternating linearized minimization method to solve the optimization problem with low-rank and sparse regularization. Learned soft singular value thresholding is introduced to ensure the clear separation of the L component and S component. Then, the iterative steps are unrolled into a network in which the regularization parameters are learnable. We prove that the proposed L+S-Net achieves global convergence under two standard assumptions. Experiments on retrospective and prospective cardiac cine datasets show that the proposed model outperforms state-of-the-art CS and existing deep learning methods and has great potential for extremely high acceleration factors (up to 24x). (C) 2021 Elsevier B.V. All rights reserved.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available