Highly undersampled magnetic resonance imaging reconstruction using autoencoding priors

• Published in: Magnetic resonance in medicine Vol. 83; no. 1; pp. 322 - 336
• Main Authors: Liu, Qiegen, Yang, Qingxin, Cheng, Huitao, Wang, Shanshan, Zhang, Minghui, Liang, Dong
• Format: Journal Article
• Language: English
• Published: United States Wiley Subscription Services, Inc 01.01.2020
• Subjects: Acceleration; Algorithms; autoencoding priors; Brain - diagnostic imaging; Computer Systems; Humans; Image processing; Image Processing, Computer-Assisted - methods; Image reconstruction; Image restoration; Machine learning; Magnetic Resonance Imaging; multichannel prior; Neural Networks, Computer; Noise generation; Noise reduction; proximal gradient descent; Sampling; Signal-To-Noise Ratio; Software; Trajectories; multichannel prior; autoencoding priors; magnetic resonance imaging; image reconstruction; proximal gradient descent
• ISSN: 0740-3194; 1522-2594; 1522-2594
• DOI: 10.1002/mrm.27921

Purpose Although recent deep learning methodologies have shown promising results in fast MR imaging, how to explore it to learn an explicit prior and leverage it into the observation constraint is still desired. Methods A denoising autoencoder (DAE) network is leveraged as an explicit prior to address the highly undersampling MR image reconstruction problem. First, inspired by the observation that the prior information learned from high‐dimension signals is more effective than that from the low‐dimension counterpart in image restoration tasks, we train the network in a multichannel scenario and apply the learned network to single‐channel image reconstruction by a variables augmentation technique. Second, because of the fact that multiple implementations of artificial noise generation in DAE favors a better underlying result, we introduce a 2‐sigma rule to complement each other for improving the final reconstruction. The whole algorithm is tackled by proximal gradient descent. Results Experimental results under varying sampling trajectories and acceleration factors consistently demonstrate the superiority of the enhanced autoencoding priors, in terms of peak signal‐to‐noise ratio, structural similarity, and high‐frequency error norm. Conclusion A simple and effective way to incorporate the DAE prior into highly undersampling MR reconstruction is proposed. Once the DAE prior is obtained, it can be applied to the reconstruction tasks with different sampling trajectories and acceleration factors, and achieves superior performance in comparison with state‐of‐the‐art methods.