Abstract:Deep learning methods for fast MRI have shown promise in reconstructing high-quality images from undersampled multi-coil k -space data, leading to reduced scan duration. However, existing methods encounter challenges related to limited receptive fields in dual-domain ( k -space and image domains) reconstruction networks, rigid data consistency operations, and suboptimal refinement structures, which collectively restrict overall reconstruction performance. This study introduces a comprehensive framework that addresses these challenges and enhances MR image reconstruction quality. Firstly, we propose Faster Inverse Fourier Convolution (FasterIFC), a frequency domain convolutional operator that significantly expands the receptive field of k -space domain reconstruction networks. Expanding the information extraction range to the entire frequency spectrum according to the spectral convolution theorem in Fourier theory enables the network to easily utilize richer redundant long-range information from adjacent, symmetrical, and diagonal locations of multi-coil k -space data. Secondly, we introduce a novel softer Data Consistency (softerDC) layer, which achieves an enhanced balance between data consistency and smoothness. This layer facilitates the implementation of diverse data consistency strategies across distinct frequency positions, addressing the inflexibility observed in current methods. Finally, we present the Dual-Domain Faster Fourier Convolution Based Network (D2F2), which features a centrosymmetric dual-domain parallel structure based on FasterIFC. This architecture optimally leverages dual-domain data characteristics while substantially expanding the receptive field in both domains. Coupled with the softerDC layer, D2F2 demonstrates superior performance on the NYU fastMRI dataset at multiple acceleration factors, surpassing state-of-the-art methods in both quantitative and qualitative evaluations.

Fourier Convolution Block with global receptive field for MRI reconstruction

Image Reconstruction for Accelerated MR Scan with Faster Fourier Convolutional Neural Networks

Dual-domain faster Fourier convolution based network for MR image reconstruction

DCT-net: Dual-domain cross-fusion transformer network for MRI reconstruction

Frequency Learning Via Multi-Scale Fourier Transformer for MRI Reconstruction.

A feature-based convolutional neural network for reconstruction of interventional MRI

HFGN: High-Frequency Residual Feature Guided Network for Fast MRI Reconstruction

CDF-Net: Cross-Domain Fusion Network for Accelerated MRI Reconstruction

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

HFF-Net: A High-Frequency Fidelity Model for Accelerated Parallel MRI Reconstruction

Deep Learning-based MRI Reconstruction with Artificial Fourier Transform (AFT)-Net

Deep magnetic resonance fingerprinting based on Local and Global Vision Transformer

Dense channel splitting network for MR image super-resolution

Isotropic Brain MRI Reconstruction from Orthogonal Scans Using 3D Convolutional Neural Network

A Hybrid Frequency-domain/Image-domain Deep Network for Magnetic Resonance Image Reconstruction

Multi-Contrast MRI Reconstruction Based on Frequency Domain Separation and Cross-Self-Attention

Accurate MRI Reconstruction Via Multi-Domain Recurrent Networks

Pyramid Convolutional RNN for MRI Image Reconstruction

Accelerated Super-resolution MR Image Reconstruction via a 3D Densely Connected Deep Convolutional Neural Network

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