FLARE-MedFM/FLARE-Task4-MRI-FM

MICCAI FLARE25

Task 4: Foundation Models for 3D CT and MRI Scans (Homepage)

This is the official dataset for MRI image foundation model development. We provide 10,000+ MRI scans for model pretraining. Downstream tasks include:

• Liver tumor segmentation: ATLAS23_liver_tumor_seg
• Cardiac tissue and pathology segmentation: EMIDEC_heart_seg_and_class
• Heart myocardial status classification (binary label: pathological vs normal): EMIDEC_heart_seg_and_class
• Autism Diagnosis: ABIDEII
• Surival Prediction: UPenn_GBM
• Neuropsychiatric Phenomics: openneuro_NeuropsychiatricPhenomics
• Brain Age Prediction: openneuro_ds004856

Dataset

Dataset Name	Task	Images (Train/Test)	Metric	Source	License
openneuro_ds004856_regression	Brain Age Prediction	600(480/120)	MAE	ds004856	CC0 1.0
openneuro_NeuropsychiatricPhenomics_classification	Neuropsychiatric Phenomics	261(148/113)	mAP	ds000030	CC0 1.0
ABIDEII_classification	Autism Diagnosis	317(253/64)	Balanced Accuracy, AUROC	ABIDEII	CC 4.0
UPenn_GBM_regression	Surival Prediction	452(361/91)	Concordance Index	UPenn_GBM	CC 4.0
ATLAS23_liver_tumor_seg	Liver tumor segmentation	60(40/20)	DSC, NSD	ATLAS	CC BY-NC-SA 4.0
EMIDEC_heart_seg_and_class	Heart Segmentation and Classification	60(40/20)	DSC, NSD, Balanced Accuracy, AUROC	EMIDEC	CC BY-NC-SA 4.0

Folder structure

FLARE-Task4-MRI-FM/
├── README.md
├── train # training set
├── val_downstream/ # validation sets for downstream tasks
│   ├── ABIDEII
│   ├── ATLAS23_liver_tumor_seg
│   ├── EMIDEC_heart_seg_and_class
│   ├── UPenn_GBM
│   ├── openneuro_NeuropsychiatricPhenomics
│   └── openneuro_ds004856
# To avoid potential data leakage, the testing set will not be released. 

Remarks

1. During model pre-training, please don't use any of the (public) annotations. Only raw images are allowed.

References

• Lalande, A., Chen, Z., Decourselle, T., Qayyum, A., Pommier, T., Lorgis, L., de La Rosa, E., Cochet, A., Cottin, Y., Ginhac, D. and Salomon, M., 2020. Emidec: a database usable for the automatic evaluation of myocardial infarction from delayed-enhancement cardiac MRI. Data, 5(4), p.89.
• Lalande, A., Chen, Z., Pommier, T., Decourselle, T., Qayyum, A., Salomon, M., Ginhac, D., Skandarani, Y., Boucher, A., Brahim, K. and de Bruijne, M., 2022. Deep learning methods for automatic evaluation of delayed enhancement-MRI. The results of the EMIDEC challenge. Medical Image Analysis, 79, p.102428.
• Quinton, F., Popoff, R., Presles, B., Leclerc, S., Meriaudeau, F., Nodari, G., Lopez, O., Pellegrinelli, J., Chevallier, O., Ginhac, D. and Vrigneaud, J.M., 2023. A tumour and liver automatic segmentation (atlas) dataset on contrast-enhanced magnetic resonance imaging for hepatocellular carcinoma. Data, 8(5), p.79.
• Denise Park and Joseph Hennessee and Evan T. Smith and Micaela Chan and Carson Katen and Julia Bacci and Sarah Frank and Sarah Monier and Alexandra Collyer and Carol Tamminga and William Moore and Neil Rofsky and Karen Rodrigue and Kristen Kennedy and Gagan Wig (2024). The Dallas Lifespan Brain Study. OpenNeuro. [Dataset]
• Bilder, R and Poldrack, R and Cannon, T and London, E and Freimer, N and Congdon, E and Karlsgodt, K and Sabb, F (2020). UCLA Consortium for Neuropsychiatric Phenomics LA5c Study. OpenNeuro. [Dataset]
• akas, S., Sako, C., Akbari, H., Bilello, M., Sotiras, A., Shukla, G., Rudie, J. D., Flores Santamaria, N., Fathi Kazerooni, A., Pati, S., Rathore, S., Mamourian, E., Ha, S. M., Parker, W., Doshi, J., Baid, U., Bergman, M., Binder, Z. A., Verma, R., Lustig, R., Desai, A. S., Bagley, S. J., Mourelatos, Z., Morrissette, J., Watt, C. D., Brem, S., Wolf, R. L., Melhem, E. R., Nasrallah, M. P., Mohan, S., O’Rourke, D. M., Davatzikos, C. (2022). The University of Pennsylvania glioblastoma (UPenn-GBM) cohort: advanced MRI, clinical, genomics, & radiomics. In Scientific Data (Vol. 9, Issue 1).
• https://fcon_1000.projects.nitrc.org/indi/abide/abide_II.html