Abstract:Background and objective: Compressed sensing (CS) has gained increased attention in magnetic resonance imaging (MRI), leveraging its efficacy to accelerate image acquisition. Incoherence measurement and nonlinear reconstruction are the most crucial guarantees of accurate restoration. However, the loose link between measurement and reconstruction hinders the further improvement of reconstruction quality, i.e., the default sampling pattern is not adaptively tailored to the downstream reconstruction method. When single-contrast reconstruction (SCR) has been upgraded to its multi-contrast reconstruction (MCR) variant, the identical morphologic information as a priori source could be integrated into the reconstruction procedure. How to measure less and reconstruct effectively by using the shareable morphologic information of various contrast images is an attractive topic. Methods: An adaptive sampling (AS) based end-to-end framework (ASSCR or ASMCR) is proposed to address this issue, which simultaneously optimizes sampling patterns and reconstruction from undersampled data in SCR or MCR scenarios. Several deep probabilistic subsampling (DPS) modules are used in AS network to construct a sampling pattern generator. In SCR and MCR, a convolution block and a data consistency layer are iteratively applied in the reconstruction network. Specifically, the learned optimal sampling pattern output from the trained AS sub-net is used for under-sampling. Incoherence measurement for single-contrast images and the combination of sampling patterns for multi-contrast data are guided by the SCR/MCR sub-net. Results: Experiments were conducted on two single-contrast and one multi-contrast public MRI datasets. Compared with several state-of-the-art reconstruction methods, SCR results show that a learned sampling pattern brings the quality of the reconstructed image closer to the fully-sampled reference. With the addition of different contrast images, under-sampled images with higher acceleration factors could be well recovered. Conclusion: The proposed method could improve the reconstruction quality of under-sampled images by using adaptive sampling patterns and learning-based reconstruction. (c) 2022 Elsevier B.V. All rights reserved.

A Deep Framework Assembling Principled Modules for CS-MRI: Unrolling Perspective, Convergence Behaviors, and Practical Modeling

Converged Deep Framework Assembling Principled Modules for CS-MRI

Sparsity-constrained SENSE reconstruction: An efficient implementation using a fast composite splitting algorithm

A Theoretically Guaranteed Deep Optimization Framework for Robust Compressive Sensing MRI.

MRI Reconstruction with Interpretable Pixel-Wise Operations Using Reinforcement Learning

A Deep Information Sharing Network for Multi-contrast Compressed Sensing MRI Reconstruction

Learing Sampling and Reconstruction Using Bregman Iteration for CS-MRI

MD-Recon-Net: A Parallel Dual-Domain Convolutional Neural Network for Compressed Sensing MRI

Deep plug-and-play MRI reconstruction based on multiple complementary priors

Compressed Sensing MRI Reconstruction with Multiple Sparsity Constraints on Radial Sampling

Joint Optimization of Cartesian Sampling Patterns and Reconstruction for Single‐contrast and Multi‐contrast Fast Magnetic Resonance Imaging

Deep Network Cascade for Dynamic Cardiac MRI Reconstruction With Motion Feature Incorporation and the Fourier Neural Attention

Inference Block 1 Inference Block 2 Inference Block 3 Inference Block 4 Inference Block

Data-Consistent Non-Cartesian Deep Subspace Learning for Efficient Dynamic MR Image Reconstruction

Joint reconstruction framework of compressed sensing and nonlinear parallel imaging for dynamic cardiac magnetic resonance imaging

A Unified Model for Compressed Sensing MRI Across Undersampling Patterns

A Deep Error Correction Network for Compressed Sensing MRI

Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images

Undersampled CS Image Reconstruction Using Nonconvex Nonsmooth Mixed Constraints

Deep MRI Reconstruction: Unrolled Optimization Algorithms Meet Neural Networks

Compressed Sensing MRI Using a Recursive Dilated Network.