DBGAN: A dual-branch generative adversarial network for undersampled MRI reconstruction

Xianzhe Liu; Hongwei Du; Jinzhang Xu; Bensheng Qiu

doi:10.1016/j.mri.2022.03.003

DBGAN: A dual-branch generative adversarial network for undersampled MRI reconstruction

Magn Reson Imaging. 2022 Jun:89:77-91. doi: 10.1016/j.mri.2022.03.003. Epub 2022 Mar 24.

Authors

Xianzhe Liu¹, Hongwei Du², Jinzhang Xu³, Bensheng Qiu¹

Affiliations

¹ Center for Biomedical Image, University of Science and Technology of China, Hefei, Anhui 230026, China.
² Center for Biomedical Image, University of Science and Technology of China, Hefei, Anhui 230026, China. Electronic address: duhw@ustc.edu.cn.
³ School of Electrical Engineering and Automation, Hefei University of Technology, Hefei, Anhui 230009, China.

PMID: 35339616
DOI: 10.1016/j.mri.2022.03.003

Abstract

Compressed sensing magnetic resonance imaging (CS-MRI) greatly accelerates the acquisition process and yield considerable reconstructed images. Deep learning was introduced into CS-MRI to further speed up the reconstruction process and improve the image quality. Recently, generative adversarial network (GAN) using two-stage cascaded U-Net structure as generator has been proven to be effective in MRI reconstruction. However, previous cascaded structure was limited to few feature information propagation channels thus may lead to information missing. In this paper, we proposed a GAN-based model, DBGAN, for MRI reconstruction from undersampled k-space data. The model uses cross-stage skip connection (CSSC) between two end-to-end cascaded U-Net in our generator to widen the channels of feature propagation. To avoid discrepancy between training and inference, we replaced classical batch normalization (BN) with instance normalization (IN) . A stage loss is involved in the loss function to boost the training performance. In addition, a bilinear interpolation decoder branch is introduced in the generator to supplement the missing information of the deconvolution decoder. Tested under five variant patterns with four undersampling rates on different modality of MRI data, the quantitative results show that DBGAN model achieves mean improvements of 3.65 dB in peak signal-to-noise ratio (PSNR) and 0.016 in normalized mean square error (NMSE) compared with state-of-the-art GAN-based methods on T1-Weighted brain dataset from MICCAI 2013 grand challenge. The qualitative visual results show that our method can reconstruct considerable images on brain and knee MRI data from different modality. Furthermore, DBGAN is light and fast - the model parameters are fewer than half of state-of-the-art GAN-based methods and each 256 × 256 image is reconstructed in 60 milliseconds, which is suitable for real-time processing.

Keywords: Bilinear interpolation decoder; CS-MRI; CSSC; GAN; IN; Stage loss; cascaded U-Net.

Publication types

Research Support, Non-U.S. Gov't

MeSH terms

Brain / diagnostic imaging
Image Processing, Computer-Assisted* / methods
Magnetic Resonance Imaging* / methods
Research Design
Signal-To-Noise Ratio