The state-of-the-art in cardiac MRI reconstruction: Results of the CMRxRecon challenge in MICCAI 2023

作     者：Lyu, Jun Qin, Chen Wang, Shuo Wang, Fanwen Li, Yan Wang, Zi Guo, Kunyuan Ouyang, Cheng Taenzer, Michael Liu, Meng Sun, Longyu Sun, Mengting Li, Qing Shi, Zhang Hua, Sha Li, Hao Chen, Zhensen Zhang, Zhenlin Xin, Bingyu Metaxas, Dimitris N. Yiasemis, George Teuwen, Jonas Zhang, Liping Chen, Weitian Zhao, Yidong Tao, Qian Pang, Yanwei Liu, Xiaohan Razumov, Artem V. Dylov, Dmitry Dou, Quan Yan, Kang Xue, Yuyang Du, Yuning Dietlmeier, Julia Garcia-Cabrera, Carles Hemidi, Ziad Al-Haj Vogt, Nora Xu, Ziqiang Zhang, Yajing Chu, Ying-Hua Chen, Weibo Bai, Wenjia Zhuang, Xiahai Qin, Jing Wu, Lianming Yang, Guang Qu, Xiaobo Wang, He Wang, Chengyan 

作者机构：Yantai Univ Sch Comp & Control Engn Yantai Peoples R China Imperial Coll London Dept Elect & Elect Engn & I X London England Fudan Univ Digital Med Res Ctr Sch Basic Med Sci Shanghai Peoples R China Imperial Coll London Bioengn Dept & Imperial X London W12 7SL England Royal Brompton Hosp Cardiovasc Magnet Resonance Unit London SW3 6NP England Kings Coll London Sch Biomed Engn & Imaging Sci London WC2R 2LS England Shanghai Jiao Tong Univ Sch Med Ruijin Hosp Dept Radiol Shanghai Peoples R China Xiamen Univ Natl Inst Data Sci Hlth & Med Dept Elect Sci Fujian Prov Key Lab Plasma & Magnet Resonance Xiamen Peoples R China Imperial Coll London Dept Comp London SW7 2AZ England Imperial Coll London Dept Brain Sci London SW7 2AZ England Fudan Univ Shanghai Pudong Hosp Shanghai Peoples R China Fudan Univ Human Phenome Inst 825 Zhangheng Rd Shanghai 201203 Peoples R China Int Human Phenome Inst Shanghai Shanghai Peoples R China Fudan Univ Zhongshan Hosp Dept Radiol Shanghai Peoples R China Shanghai Jiao Tong Univ RuiJin Hosp Lu Wan Branch Dept Urol Sch Med Shanghai Peoples R China Fudan Univ Inst Sci & Technol Brain Inspired Intelligence Shanghai 200433 Peoples R China Rutgers State Univ Dept Comp Sci New Brunswick NJ 08901 USA Netherlands Canc Inst Dept Immunol Plesmanlaan 121 NL-1066 CX Amsterdam Netherlands Chinese Univ Hong Kong Dept Imaging & Intervent Radiol CUHK Lab AI Radiol CLAIR Shatin Hong Kong Peoples R China Delft Univ Technol Dept Imaging Phys Lorentzweg 1 NL-2628 CJ Delft Netherlands Tianjin Univ Sch Elect & Informat Engn TJK BIIT Lab Tianjin 300072 Peoples R China Inst Appl Phys & Computat Math Beijing 100094 Peoples R China Skolkovo Inst Sci & Technol Ctr Artificial Intelligence Technol Bolshoy Blvd 30Bld 1 Moscow 121205 Russia Univ Virginia Dept Biomed Engn 415 Lane Rd Charlottesville VA 22903 USA Univ Edinburgh Inst Imaging Data & Commun IDCOM Edinburgh EH9 3FG Scotland Dublin City Univ Insight SFI Res Ctr Data Analyt Dublin 9 Ireland Dublin City Univ SFI Ctr Res Training Machine Learning ML Labs Dublin Ireland Univ Lubeck Inst Med Informat Ratzeburger Allee 160 D-23562 Lubeck Germany Univ Lorraine INSERM U1254 IADI Rue Morvan F-54511 Nancy France Univ Shanghai Sci & Technol Sch Hlth Sci & Engn Shanghai Peoples R China GE Healthcare Sci & Technol Dept Beijing Peoples R China Siemens Healthineers Ltd Shanghai Peoples R China Philips Healthcare Shanghai Peoples R China Fudan Univ Sch Data Sci Shanghai Peoples R China Hong Kong Polytech Univ Sch Nursing Hong Kong Peoples R China Shanghai Jiao Tong Univ Sch Med Ren Ji Hosp Dept Radiol Shanghai 200127 Peoples R China Artificial Intelligence Res Inst Moscow 121170 Russia 

出 版 物：《MEDICAL IMAGE ANALYSIS》 (Med. Image Anal.)

年 卷 期：2025年第101卷

页      面：103485页

核心收录：

学科分类：0831[工学-生物医学工程（可授工学、理学、医学学位）] 0812[工学-计算机科学与技术（可授工学、理学学位）] 1009[医学-特种医学] 10[医学] 

基　　金：National Natural Science Foundation of China [62371413, 62331021, 62122064] Yantai Basic Research Key Project, China [2023JCYJ041] Youth Innovation Science and Tech-nology Support Program of Shandong Provincial, China [2023KJ239] Youth Program of Natural Science Founda-tion of Shandong Province [ZR2024QF001] Natural Science Foundation of Fujian Province of China [2023J02005] President Fund of Xiamen University, China EPSRC, UK [EP/X039277/1] Imperial College London, UK [EP/S023283/1] ERC IMI [101005122, 952172] MRC [MC/PC/21013] Royal Society [IEC\NSFC\ 211235] NVIDIA Academic Hardware Grant Program Boehringer Ingelheim Ltd [RDA01] Horizon Europe MSCA [EP/Z002206/1, MR/V023799/1] National Institutes of Health (NIH) , United States grant [7R01HL148788-03] Royal Academy of Engineering [RCSRF1819\8\25] UK's Engineering and Physical Sciences Research Council (EPSRC) [EP/X017680/1] China Scholarship Council Shanghai Municipal Science and Technology Major Project [2023SHZD2X02A05] Shanghai Rising-Star Program [24QA2703300] 

主　　题：Reconstruction Cardiac imaging Fast imaging Under-sampling K-space 

摘      要：Cardiac magnetic resonance imaging (MRI) provides detailed and quantitative evaluation of the heart s structure, function, and tissue characteristics with high-resolution spatial-temporal imaging. However, its slow imaging speed and motion artifacts are notable limitations. Undersampling reconstruction, especially data- driven algorithms, has emerged as a promising solution to accelerate scans and enhance imaging performance using highly under-sampled data. Nevertheless, the scarcity of publicly available cardiac k-space datasets and evaluation platform hinder the development of data-driven reconstruction algorithms. To address this issue, we organized the Cardiac MRI Reconstruction Challenge (CMRxRecon) in 2023, in collaboration with the 26th International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI). CMRxRecon presented an extensive k-space dataset comprising cine and mapping raw data, accompanied by detailed annotations of cardiac anatomical structures. With overwhelming participation, the challenge attracted more than 285 teams and over 600 participants. Among them, 22 teams successfully submitted Docker containers for the testing phase, with 7 teams submitted for both cine and mapping tasks. All teams use deep learning based approaches, indicating that deep learning has predominately become a promising solution for the problem. The first-place winner of both tasks utilizes the E2E-VarNet architecture as backbones. In contrast, U-Net is still the most popular backbone for both multi-coil and single-coil reconstructions. This paper provides a comprehensive overview of the challenge design, presents a summary of the submitted results, reviews the employed methods, and offers an in-depth discussion that aims to inspire future advancements in cardiac MRI reconstruction models. The summary emphasizes the effective strategies observed in Cardiac MRI reconstruction, including backbone architecture, loss function, pre-processing tech