CARE-Myocardium

Myocardial Pathology Segmentation

Registration

To access the dataset, please register here.

Motivation

Figure 1. Myocardial pathology segmentation and its challenges. (A) Myocardial Pathology Segmentation: Scar and edema regions are marked in green and yellow, respectively. (B) Cine MyoPS: Scar regions are marked in blue. (C) Challenges of Myocardial Pathology Segmentation: The challenges include multi-center data, missing sequences, and misalignments in multi-sequence CMR images.

Myocardial infarction (MI) is a leading cause of mortality and disability worldwide. Accurate assessment of myocardial viability is essential for the diagnosis and clinical management of MI patients. Multi-sequence cardiac magnetic resonance (MS-CMR) imaging provides comprehensive insights into myocardial pathology, playing a pivotal role in tailored patient care. As shown in Figure 1 (A), balanced steady-state free precession (bSSFP) cine sequences clearly delineate anatomical boundaries, while late gadolinium enhancement (LGE) and T2-weighted (T2) CMR sequences visualize myocardial scars and edema, respectively.

Despite the high efficacy of LGE in visualizing myocardial scars, its acquisition can be time-consuming and sometimes clinically prohibitive due to the required administration of contrast agents, which highlights an urgent need for contrast-free alternatives. Cine CMR emerges as a rapid, contrast-free imaging technique capable of capturing myocardial dynamics. As shown in Figure 1 (B), non-viable myocardial regions (scars) exhibit distinct abnormalities in cardiac motion and wall thickness compared to viable tissue. Leveraging these functional differences, cine CMR presents a promising yet highly challenging alternative for direct myocardial pathology segmentation.

Tasks

The primary objective of this track is to accurately segment myocardial pathology regions (specifically, scars and edema) from CMR sequences. This track seeks innovative solutions to address two distinct subtasks: MyoPS, which focuses on segmenting scars and edema using MS-CMR data, and CineMyoPS, which focuses on segmenting scars using solely cine CMR data. We encourage participants to develop robust algorithms capable of overcoming significant real-world challenges, including multi-center data variations, missing sequences in certain centers, and spatial misalignments across multi-sequence CMRs, as illustrated in Figure 1 (C).

MyoPS data: LGE, T2 and bSSFP sequences. The segmentation target is scars and edema at the LGE and T2 sequences, respectively.
CineMyoPS data: cine sequence. The segmentation target is scars at the end-diastolic frame of the cine sequence.

The specific substructures, each associated with a unique label value, are:

Scar - Label value: 2221
Edema - Label value: 1220
Left ventricle - Label value: 500
Myocardium - Label value: 200
Right ventricle - Label value: 600

We will rank participant methods based on the settings (Lb1–Lb3) detailed in the following table:

Leaderboard (Lb) for Myocardium track across targets and evaluation settings. Lb1–Lb3 represent performance from different subtasks and targets (e.g., Lb1 = Scar segmentation peformance on the MyoPS subtask).
Subtask	Target	Leaderboard
MyoPS	Scar	Lb1
MyoPS	Edema	Lb2
CineMyoPS	Scar	Lb3

Data details

Training data – MyoPS

Center	Num. patients	Sequences	Manual labels
A	81	LGE	Scar, left ventricle and myocardium
B	35	LGE, T2 and bSSFP	Scar, edema, left ventricle, myocardium and right ventricle
C	45	LGE, T2 and bSSFP	Scar, edema, left ventricle, myocardium and right ventricle
E	7	LGE and bSSFP	Scar, left ventricle, myocardium and right ventricle
F	9	LGE and bSSFP	Scar, left ventricle, myocardium and and right ventricle
G	8	LGE and bSSFP	Scar, left ventricle, myocardium and and right ventricle
H	35	LGE	Scar

Validation data – MyoPS

Center	Num. patients	Sequences	Manual labels
(Anonymous)	15	LGE, T2 and bSSFP	Scar, edema, left ventricle, myocardium and right ventricle

Test data – MyoPS

Center	Num. patients	Sequences	Manual labels
(Anonymous)	65	LGE, T2 and bSSFP	Scar, edema, left ventricle, myocardium and right ventricle

Training data – CineMyoPS

Center	Num. patients	Sequences	Manual labels
α	40	Cine	Scar, left ventricle and myocardium
β	24	Cine	Scar, left ventricle and myocardium

Validation data – CineMyoPS

Center	Num. patients	Sequences	Manual labels
(Anonymous)	15	Cine	Scar, left ventricle and myocardium

Test data – CineMyoPS

Center	Num. patients	Sequences	Manual labels
(Anonymous)	45	Cine	Scar, left ventricle and myocardium

Metrics

The performance of scar and edema segmentation results will be evaluated by：

Dice Similarity Coefficient (Dice)
Hausdorff Distance (HD in mm)

Rules

Only automatic methods are acceptable. Participants must utilize algorithms that do not require manual intervention or human-assisted processes for the segmentation task.
Pre-trained models are allowed in this track. The solutions could be developed with pre-trained fundation models, such as SAM, CLIP, and MedSAM.

Citations

Please cite these papers when you use the data for publications:

 @article{zhuang2019multivariate,
    title={Multivariate mixture model for myocardial segmentation combining multi-source images},
    author={Zhuang, Xiahai},
    journal={IEEE transactions on pattern analysis and machine intelligence},
    volume={41},
    number={12},
    pages={2933--2946},
    year={2019},
}

@article{qiu2023myops,
  title={MyoPS-Net: Myocardial pathology segmentation with flexible combination of multi-sequence CMR images},
  author={Qiu, Junyi and Li, Lei and Wang, Sihan and Zhang, Ke and Chen, Yinyin and Yang, Shan and Zhuang, Xiahai},
  journal={Medical image analysis},
  volume={84},
  pages={102694},
  year={2023},
}

 @article{ding2023aligning,
  title={Aligning multi-sequence CMR towards fully automated myocardial pathology segmentation},
  author={Ding, Wangbin and Li, Lei and Qiu, Junyi and Wang, Sihan and Huang, Liqin and Chen, Yinyin and Yang, Shan and Zhuang, Xiahai},
  journal={IEEE Transactions on Medical Imaging},
  year={2023},
}

@article{ding2025cinemyops,
  title={CineMyoPS: Segmenting Myocardial Pathologies from Cine Cardiac MR},
  author={Ding, Wangbin and Li, Lei and Qiu, Junyi and Lin, Bogen and Yang, Mingjing and Huang, Liqin and Wu, Lianming and Wang, Sihan and Zhuang, Xiahai},
  journal={IEEE Transactions on Medical Imaging},
  year={2025},
}