Join the challenge!

1. Sign up on the Synapse website.
2. Please fill out the registration form, including your team name, team members, Synaspe account, and sign the data agreement(pdf). 

Train the model

Participants are expected to train models in their local computational environments and to submit docker containers on Synapse platform. A leaderboard will be maintained on the Synapse platform during the validation phase.

Download the data

Once we received the registration form and the data agreement  from your team, you have the access to download the data.
[1 Apr, 2024] Training data and validation data will be released. The results can be evaluated from 1 Aug, 2024.
Please follow the instructions on Synapse to download the data.
We will send you the link and password to download the data (via Google Drive , One Drive, and  Baidu Netdisk)

Code availability

We provide the code to facilitate the use of the data we release at https://github.com/CmrxRecon/CMRxRecon.
A brief description of the provided package is as follows: 

Evaluation platform

Validation of the received docker on unseen test set will be performed on a cloud server with a configuration as follows:

• CPU: Intel(R) Core(TM) i9-10900K CPU @ 3.70GHz, 20 cores; 
• RAM: 128 GB; GPU: NVIDIA GeForce RTX 3090 (24 GB VRAM, single GPU); 
• GPU DRIVER: NVIDIA-DRIVER Version 515.48.07; 
• GPU CUDA: Version 11.7;
• Validation Cases: 50 cases;
• Time Limitation: 4 hours / team for each task 

Publication references

You are free to use and/or refer to the CMRxRecon challenge and datasets in your own research after the embargo period (Dec 2023), provided that you cite the following manuscripts: 

Reference of the imaging acquisition protocol:
1.Wang C, Lyu J, Wang S, et al. CMRxRecon: An open cardiac MRI dataset for the competition of accelerated image reconstruction[J]. arXiv preprint arXiv:2309.10836, 2023.
2.Wang C, Li Y, Lv J, et al. Recommendation for Cardiac Magnetic Resonance Imaging-Based Phenotypic Study: Imaging Part. Phenomics. 2021, 1(4): 151-170.https://doi.org/10.1007/s43657-021-00018-x

Other reference (optional for citation):
1.Wang C, Jang J, Neisius U, et al. Black blood myocardial T2 mapping. Magnetic resonance in medicine. 2019, 81(1): 153-166. https://doi.org/10.1002/mrm.27360
2.Lyu J, Li G, Wang C, et al. Region-focused multi-view transformer-based generative adversarial network for cardiac cine MRI reconstruction[J]. Medical Image Analysis, 2023: 102760. https://doi.org/10.1016/j.media.2023.102760
3.Qin C, Schlemper J, Caballero J, et al. Convolutional recurrent neural networks for dynamic MR image reconstruction. IEEE transactions on medical imaging, 2018, 38(1): 280-290. https://doi.org/10.1109/TMI.2018.2863670
4.Qin C, Duan J, Hammernik K, et al. Complementary time‐frequency domain networks for dynamic parallel MR image reconstruction. Magnetic Resonance in Medicine, 2021, 86(6): 3274-3291. https://doi.org/10.1002/mrm.28917
5.Lyu J, Tong X, Wang C. Parallel Imaging With a Combination of SENSE and Generative Adversarial Networks (GAN). Quantitative Imaging in Medicine and Surgery. 2020, 10(12): 2260–2273. https://doi.org/10.21037/qims-20-518
6.Lyu J, Sui B, Wang C, et al. DuDoCAF: Dual-Domain Cross-Attention Fusion with Recurrent Transformer for Fast Multi-contrast MR Imaging. International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, Cham, 2022: 474-484.
7.Wang S, Qin C, Wang C, et al. The Extreme Cardiac MRI Analysis Challenge under Respiratory Motion (CMRxMotion). arXiv preprint arXiv:2210.06385, 2022.
8.Shangqi Gao, Hangqi Zhou, Yibo Gao, Xiahai Zhuang. BayeSeg: Bayesian Modeling for Medical Image Segmentation with Interpretable Generalizability. Medical Image Analysis Volume 89, 102889, 2023 (Elsevier-MedIA 1st Prize & MICCAl Best Paper Award 2023)
9.Wang Z, Qian C, Guo D, Sun H, Li R, Zhao B, Qu X, One-dimensional Deep Low-rank and Sparse Network for Accelerated MRI, IEEE Transactions on Medical Imaging, 42: 79-90, 2023. https://doi.org/10.1109/TMI.2022.3203312
10.Wang Z, et al., Deep Separable Spatiotemporal Learning for Fast Dynamic Cardiac MRI, arXiv preprint arXiv:2402.15939, 2024. https://arxiv.org/abs/2402.15939

Principal of participation

Note: Participants are not required to upload the complete training code. But teams willing to upload the original training code will be automatically entered into the code-sharing pool.

Awards

We will provide monetary awards for the top 5 winners of each task. The prize pool is exclusively sponsored by Philips.

Task 1: Multi-contrast CMR reconstruction

Task 2: Random sampling CMR reconstruction

All submissions will be reported in the leaderboard. Each participating team can participate in both tasks. However, we only present the higher reward among the two tasks to each team.

Prize-winning methods will be announced publicly as part of a scientific session at the MICCAI annual meeting.