Deep learning based MRI reconstruction with transformer

• Published in: Computer methods and programs in biomedicine 2023-05, Vol.233, p.107452-107452, Article 107452
• Main Authors: Wu, Zhengliang, Liao, Weibin, Yan, Chao, Zhao, Mangsuo, Liu, Guowen, Ma, Ning, Li, Xuesong
• Format: Article
• Language: English
• Subjects: Algorithms; Compress sensing; Deep Learning; Image Processing, Computer-Assisted - methods; Magnetic resonance imaging (MRI); Magnetic Resonance Imaging - methods; Neural Networks, Computer; Transformer
• ISSN: 0169-2607; 1872-7565
• DOI: 10.1016/j.cmpb.2023.107452

•New approach leveraging Transformer’s powerful processing ability to recover structure in deep MRI reconstruction.•Proposed models achieve best reconstruction performances in reconstruction tasks of different contrasts and anatomies.•K-space consistency layers are combined in the model to reach local optimum and get better fidelity.•Visually coherent results can be generated in extremely low sampling ratio with 10 times speed-up. Magnetic resonance imaging (MRI) has become one of the most powerful imaging techniques in medical diagnosis, yet the prolonged scanning time becomes a bottleneck for application. Reconstruction methods based on compress sensing (CS) have made progress in reducing this cost by acquiring fewer points in the k-space. Traditional CS methods impose restrictions from different sparse domains to regularize the optimization that always requires balancing time with accuracy. Neural network techniques enable learning a better prior from sample pairs and generating the results in an analytic way. In this paper, we propose a deep learning based reconstruction method to restore high-quality MRI images from undersampled k-space data in an end-to-end style. Unlike prior literature adopting convolutional neural networks (CNN), advanced Swin Transformer is used as the backbone of our work, which proved to be powerful in extracting deep features of the image. In addition, we combined the k-space consistency in the output and further improved the quality. We compared our models with several reconstruction methods and variants, and the experiment results proved that our model achieves the best results in samples at low sampling rates. The source code of KTMR could be acquired at https://github.com/BITwzl/KTMR.