DeepcomplexMRI: Exploiting deep residual network for fast parallel MR imaging with complex convolution

• Published in: Magnetic resonance imaging Vol. 68; pp. 136 - 147
• Main Authors: Wang, Shanshan, Cheng, Huitao, Ying, Leslie, Xiao, Taohui, Ke, Ziwen, Zheng, Hairong, Liang, Dong
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier Inc 01.05.2020
• Subjects: Convolutional neural network; Deep learning; Fast MR imaging; Parallel imaging; Prior knowledge; Deep learning; Prior knowledge; Parallel imaging; Convolutional neural network; Fast MR imaging
• ISSN: 0730-725X; 1873-5894; 1873-5894
• DOI: 10.1016/j.mri.2020.02.002

This paper proposes a multi-channel image reconstruction method, named DeepcomplexMRI, to accelerate parallel MR imaging with residual complex convolutional neural network. Different from most existing works which rely on the utilization of the coil sensitivities or prior information of predefined transforms, DeepcomplexMRI takes advantage of the availability of a large number of existing multi-channel groudtruth images and uses them as target data to train the deep residual convolutional neural network offline. In particular, a complex convolutional network is proposed to take into account the correlation between the real and imaginary parts of MR images. In addition, the k-space data consistency is further enforced repeatedly in between layers of the network. The evaluations on in vivo datasets show that the proposed method has the capability to recover the desired multi-channel images. Its comparison with state-of-the-art methods also demonstrates that the proposed method can reconstruct the desired MR images more accurately. •Proposed an end-to-end parallel imaging reconstruction framework. The proposed framework doesn’t need any calculation of the sensitivity information to resolve the aliasing and correlations among the channels.•Both real-valued and complex-valued versions of our proposed framework have been investigated for parallel imaging with our code released to the public.•The method has been compared withquite a few state-of-the-art methods. Encouraging performances have been achieved by our proposed framework.