Abstract:Background: Compressed sensing has been employed to accelerate magnetic resonance imaging by sampling fewer measurements. However, conventional iterative optimization-based CS-MRI are time-consuming for iterative calculations and often share poor generalization ability on multicontrast datasets. Most deep-learning-based CS-MRI focus on learning an end-to-end mapping while ignoring some prior knowledge existed in MR images. Purpose: We propose an iterative fusion model to integrate the image and gradient-based priors into reconstruction via convolutional neural network models while maintaining high quality and preserving the detailed information as well. Methods: We propose a gradient-enhanced fusion network (GFN) for fast and accurate MRI reconstruction, in which dense blocks with dilated convolution and dense residual learnings are used to capture abundant features with fewer parameters. Meanwhile, decomposed gradient maps containing obvious structural information are introduced into the network to enhance the reconstructed images. Besides this, gradient-based priors along directions X and Y are exploited to learn adaptive tight frames for reconstructing the desired image contrast and edges by respective gradient fusion networks. After that, both image and gradient priors are fused in the proposed optimization model, in which we employ the l2 -norm to promote the sparsity of gradient priors. The proposed fusion model can effectively help to capture edge structures in the gradient images and to preserve more detailed information of MR images. Results: Experimental results demonstrate that the proposed method outperforms several CS-MRI methods in terms of peak signal-to-noise (PSNR), the structural similarity index (SSIM), and visualizations on three sampling masks with different rates. Noteworthy, to evaluate the generalization ability, the proposed model conducts cross-center training and testing experiments for all three modalities and shares more stable performance compared than other approaches. In addition, the proposed fusion model is applied to other comparable deep learning methods. The quantitative results show that the reconstruction results of these methods are obviously improved. Conclusions: The gradient-based priors reconstructed from GFNs can effectively enhance the edges and details of under-sampled data. The proposed fusion model integrates image and gradient priors using l2 -norm can effectively improve the generalization ability on multicontrast datasets reconstruction.

Compressed Sensing MRI by Integrating Deep Denoiser and Weighted Schatten P-Norm Minimization

Accelerating CS-MRI Reconstruction with Fine-Tuning Wasserstein Generative Adversarial Network

EiCSNet: Efficient Iterative Neural Network for Compressed Sensing Reconstruction

Single Image Compressed Sensing MRI via a Self-Supervised Deep Denoising Approach

Compressed Sensing MRI Reconstruction with Multiple Sparsity Constraints on Radial Sampling

Deep Convolutional Neural Network for Compressive Sensing of Magnetic Resonance Images

A Deep Ensemble Network for Compressed Sensing MRI

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

MS-DRDNet: Optimization-Inspired Deep Compressive Sensing Network for MRI

Compressed Sensing Undersampled MRI Reconstruction Using Iterative Shrinkage Thresholding Based on NSST

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

Compressed Sensing MR Image Reconstruction Via a Deep Frequency-Division Network.

CS-MRI Reconstruction Via Group-Based Eigenvalue Decomposition and Estimation

Nonlocal low-rank plus deep denoising prior for robust image compressed sensing reconstruction

High-dimensional embedding network derived prior for compressive sensing MRI reconstruction

Compressed Sensing MRI via a Multi-scale Dilated Residual Convolution Network

A Unified Model for Compressed Sensing MRI Across Undersampling Patterns

Deep compressed sensing MRI via a gradient-enhanced fusion model

Compressed Sensing Mri Reconstruction from Highly Undersampled K-Space Data Using Nonsubsampled Shearlet Transform Sparsity Prior

Deep ADMM-Net for Compressive Sensing MRI