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

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.

References in corpus (4)

• Deep Learning in Neural Networks: An Overview
• Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
• Deep Unfolding: Model-Based Inspiration of Novel Deep Architectures
• Maximal Sparsity with Deep Networks?

Cited by in corpus (21)

• fastMRI: An Open Dataset and Benchmarks for Accelerated MRI
• Model-Driven Deep Learning for MIMO Detection
• Advancing machine learning for MR image reconstruction with an open competition: Overview of the 2019 fastMRI challenge
• A Review on Deep Learning in Medical Image Reconstruction
• InDuDoNet+: A Deep Unfolding Dual Domain Network for Metal Artifact Reduction in CT Images
• Multispectral and Hyperspectral Image Fusion by MS/HS Fusion Net
• $\Sigma$-net: Systematic Evaluation of Iterative Deep Neural Networks for Fast Parallel MR Image Reconstruction
• ADMM-DAD net: a deep unfolding network for analysis compressed sensing
• Assessment of Data Consistency through Cascades of Independently Recurrent Inference Machines for fast and robust accelerated MRI reconstruction
• Deep Learning for Biomedical Image Reconstruction: A Survey
• RCDNet: An Interpretable Rain Convolutional Dictionary Network for Single Image Deraining
• Compressed Sensing: From Research to Clinical Practice with Data-Driven Learning
• On the Convergence of Learning-based Iterative Methods for Nonconvex Inverse Problems
• Deep Learning Techniques for Compressive Sensing-Based Reconstruction and Inference -- A Ubiquitous Systems Perspective
• MetaInv-Net: Meta Inversion Network for Sparse View CT Image Reconstruction
• Understanding Deep Architectures with Reasoning Layer
• Investigating Customization Strategies and Convergence Behaviors of Task-specific ADMM
• Compressed Sensing MRI via a Multi-scale Dilated Residual Convolution Network
• SEGAN: Structure-Enhanced Generative Adversarial Network for Compressed Sensing MRI Reconstruction
• Tree Search Network for Sparse Regression
• Outlier Detection Using Generative Models with Theoretical Performance Guarantees