ADMM-Net: A Deep Learning Approach for Compressive Sensing MRI

TL;DR

This paper introduces ADMM-Nets, deep learning architectures inspired by ADMM algorithms, for fast and accurate MRI image reconstruction from under-sampled data, outperforming previous methods.

Contribution

The paper presents novel deep architectures based on ADMM for MRI reconstruction, extending to complex images and trained end-to-end for improved accuracy and speed.

Findings

Achieves state-of-the-art reconstruction accuracy

Operates with computational speed comparable to ADMM

Effective on various sampling rates

Abstract

Compressive sensing (CS) is an effective approach for fast Magnetic Resonance Imaging (MRI). It aims at reconstructing MR images from a small number of under-sampled data in k-space, and accelerating the data acquisition in MRI. To improve the current MRI system in reconstruction accuracy and speed, in this paper, we propose two novel deep architectures, dubbed ADMM-Nets in basic and generalized versions. ADMM-Nets are defined over data flow graphs, which are derived from the iterative procedures in Alternating Direction Method of Multipliers (ADMM) algorithm for optimizing a general CS-based MRI model. They take the sampled k-space data as inputs and output reconstructed MR images. Moreover, we extend our network to cope with complex-valued MR images. In the training phase, all parameters of the nets, e.g., transforms, shrinkage functions, etc., are discriminatively trained end-to-end. In the testing phase, they have computational overhead similar to ADMM algorithm but use optimized parameters learned from the data for CS-based reconstruction task. We investigate different configurations in network structures and conduct extensive experiments on MR image reconstruction under different sampling rates. Due to the combination of the advantages in model-based approach and deep learning approach, the ADMM-Nets achieve state-of-the-art reconstruction accuracies with fast computational speed.