Towards Modality- and Sampling-Universal Learning Strategies for Accelerating Cardiovascular Imaging: Summary of the CMRxRecon2024 Challenge
Abstract
Cardiovascular health is vital to human well-being, and cardiac magnetic resonance (CMR) imaging is considered the {clinical reference standard} for diagnosing cardiovascular disease. However, its adoption is hindered by long scan times, complex contrasts, and inconsistent quality. While deep learning methods perform well on specific CMR imaging {sequences}, they often fail to generalize across modalities and sampling schemes. The lack of benchmarks for high-quality, fast CMR image reconstruction further limits technology comparison and adoption. The CMRxRecon2024 challenge, attracting over 200 teams from 18 countries, addressed these issues with two tasks: generalization to unseen {modalities} and robustness to diverse undersampling patterns. We introduced the largest public multi-{modality} CMR raw dataset, an open benchmarking platform, and shared code. Analysis of the best-performing solutions revealed that prompt-based adaptation and enhanced physics-driven consistency enabled strong cross-scenario performance. These findings establish principles for generalizable reconstruction models and advance clinically translatable AI in cardiovascular imaging.
Cite
@article{arxiv.2503.03971,
title = {Towards Modality- and Sampling-Universal Learning Strategies for Accelerating Cardiovascular Imaging: Summary of the CMRxRecon2024 Challenge},
author = {Fanwen Wang and Zi Wang and Yan Li and Jun Lyu and Chen Qin and Shuo Wang and Kunyuan Guo and Mengting Sun and Mingkai Huang and Haoyu Zhang and Michael Tänzer and Qirong Li and Xinran Chen and Jiahao Huang and Yinzhe Wu and Haosen Zhang and Kian Anvari Hamedani and Yuntong Lyu and Longyu Sun and Qing Li and Tianxing He and Lizhen Lan and Qiong Yao and Ziqiang Xu and Bingyu Xin and Dimitris N. Metaxas and Narges Razizadeh and Shahabedin Nabavi and George Yiasemis and Jonas Teuwen and Zhenxi Zhang and Sha Wang and Chi Zhang and Daniel B. Ennis and Zhihao Xue and Chenxi Hu and Ruru Xu and Ilkay Oksuz and Donghang Lyu and Yanxin Huang and Xinrui Guo and Ruqian Hao and Jaykumar H. Patel and Guanke Cai and Binghua Chen and Yajing Zhang and Sha Hua and Zhensen Chen and Qi Dou and Xiahai Zhuang and Qian Tao and Wenjia Bai and Jing Qin and He Wang and Claudia Prieto and Michael Markl and Alistair Young and Hao Li and Xihong Hu and Lianming Wu and Xiaobo Qu and Guang Yang and Chengyan Wang},
journal= {arXiv preprint arXiv:2503.03971},
year = {2025}
}
Comments
15 pages