# GENRE-CMR: Generalizable Deep Learning for Diverse Multi-Domain Cardiac MRI Reconstruction

arXiv: https://arxiv.org/abs/2508.20600 · HTML: https://mlsanity.com/paper/2508.20600/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 16 of its 31 references that are in the map.
- First posted on arXiv 28 Aug 2025.

## Builds on (most cited first, 16 of 16)

- [Domain Adaptation for Medical Image Analysis: A Survey](https://mlsanity.com/paper/2102.09508/) (2021)
- [Multi-Loss Weighting with Coefficient of Variations](https://mlsanity.com/paper/2009.01717/) (2020)
- [Fill the K-Space and Refine the Image: Prompting for Dynamic and Multi-Contrast MRI Reconstruction](https://mlsanity.com/paper/2309.13839/) (2023)
- [The state-of-the-art in Cardiac MRI Reconstruction: Results of the CMRxRecon Challenge in MICCAI 2023](https://mlsanity.com/paper/2404.01082/) (2024)
- [Clean self-supervised MRI reconstruction from noisy, sub-sampled training data with Robust SSDU](https://mlsanity.com/paper/2210.01696/) (2022)
- [Towards Modality- and Sampling-Universal Learning Strategies for Accelerating Cardiovascular Imaging: Summary of the CMRxRecon2024 Challenge](https://mlsanity.com/paper/2503.03971/) (2025)
- [Generalizing Deep Learning MRI Reconstruction across Different Domains](https://mlsanity.com/paper/1902.10815/) (2019)
- [A Generalised Deep Meta-Learning Model for Automated Quality Control of Cardiovascular Magnetic Resonance Images](https://mlsanity.com/paper/2303.13324/) (2023)
- [MRPD: Undersampled MRI reconstruction by prompting a large latent diffusion model](https://mlsanity.com/paper/2402.10609/) (2024)
- [Statistical Distance-Guided Unsupervised Domain Adaptation for Automated Multi-Class Cardiovascular Magnetic Resonance Image Quality Assessment](https://mlsanity.com/paper/2409.00375/) (2024)
- [HyperCMR: Enhanced Multi-Contrast CMR Reconstruction with Eagle Loss](https://mlsanity.com/paper/2410.03624/) (2024)
- [Deep Multi-contrast Cardiac MRI Reconstruction via vSHARP with Auxiliary Refinement Network](https://mlsanity.com/paper/2411.01291/) (2024)
- [An All-in-one Approach for Accelerated Cardiac MRI Reconstruction](https://mlsanity.com/paper/2411.10787/) (2024)
- [FractMorph: A Fractional Fourier-Based Multi-Domain Transformer for Deformable Image Registration](https://mlsanity.com/paper/2508.12445/) (2025)
- [Multiple Teachers-Meticulous Student: A Domain Adaptive Meta-Knowledge Distillation Model for Medical Image Classification](https://mlsanity.com/paper/2403.11226/) (2024)
- [Low-Rank Conjugate Gradient-Net for Accelerated Cardiac MR Imaging](https://mlsanity.com/paper/2411.11175/) (2024)

## Built on by (most cited first, 2 of 2)

- [SDUM: A Scalable Deep Unrolled Model for Universal MRI Reconstruction](https://mlsanity.com/paper/2512.17137/) (2025)
- [Joint Supervised and Self-Supervised Training with Acquisition-Robust Techniques for Accelerated 4D Flow MRI Reconstruction](https://mlsanity.com/paper/2609.38644/) (2026)

## Abstract (arXiv)

Accelerated Cardiovascular Magnetic Resonance (CMR) image reconstruction remains a critical challenge due to the trade-off between scan time and image quality, particularly when generalizing across diverse acquisition settings. We propose GENRE-CMR, a generative adversarial network (GAN)-based architecture employing a residual deep unrolled reconstruction framework to enhance reconstruction fidelity and generalization. The architecture unrolls iterative optimization into a cascade of convolutional subnetworks, enriched with residual connections to enable progressive feature propagation from shallow to deeper stages. To further improve performance, we integrate two loss functions: (1) an Edge-Aware Region (EAR) loss, which guides the network to focus on structurally informative regions and helps prevent common reconstruction blurriness; and (2) a Statistical Distribution Alignment (SDA) loss, which regularizes the feature space across diverse data distributions via a symmetric KL divergence formulation. Extensive experiments confirm that GENRE-CMR surpasses state-of-the-art methods on training and unseen data, achieving 0.9552 SSIM and 38.90 dB PSNR on unseen distributions across various acceleration factors and sampling trajectories. Ablation studies confirm the contribution of each proposed component to reconstruction quality and generalization. Our framework presents a unified and robust solution for high-quality CMR reconstruction, paving the way for clinically adaptable deployment across heterogeneous acquisition protocols.