DIMENSION: Dynamic MR imaging with both k‐space and spatial prior knowledge obtained via multi‐supervised network training. (4th September 2019)

• Record Type: Journal Article
• Authors: Wang, Shanshan; Ke, Ziwen; Cheng, Huitao; Jia, Sen; Ying, Leslie; Zheng, Hairong; Liang, Dong
• Other Names: Zhang Hui guestEditor.; Alexander Daniel C. guestEditor.; Shen Dinggang guestEditor.; Yap Pew‐Thian guestEditor.
• Is Part Of: NMR in biomedicine. Volume 35:Number 4(2022)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2019-09-04
• Subjects: compressed sensing -- deep learning -- dynamic MR imaging -- k‐space prior -- multi‐supervised; Nuclear magnetic resonance -- Periodicals; Magnetic Resonance Spectroscopy -- Periodicals; 574
• DOI: 10.1002/nbm.4131
• Languages: English
• ISSNs: 0952-3480

Abstract

Abstract : Dynamic MR image reconstruction from incomplete k‐space data has generated great research interest due to its capability in reducing scan time. Nevertheless, the reconstruction problem is still challenging due to its ill‐posed nature. Most existing methods either suffer from long iterative reconstruction time or explore limited prior knowledge. This paper proposes a dynamic MR imaging method with both k‐space and spatial prior knowledge integrated via multi‐supervised network training, dubbed as DIMENSION. Specifically, the DIMENSION architecture consists of a frequential prior network for updating the k‐space with its network prediction and a spatial prior network for capturing image structures and details. Furthermore, a multi‐supervised network training technique is developed to constrain the frequency domain information and the spatial domain information. The comparisons with classical k‐t FOCUSS, k‐t SLR, L+S and the state‐of‐the‐art CNN‐based method on in vivo datasets show our method can achieve improved reconstruction results in shorter time. Abstract : The graphical abstract of the proposed DIMENSION method. The network architecture consists of a frequential domain network (FDN) for updating the k‐space and a spatial domain network (SDN) for capturing image structures and details. A multi‐supervised network training technique is developed, where the k‐space loss and the spatial loss are proposed to constrain the frequency domain information and reconstruction results at different levels.