EVALUATION
Piooneering generalizability and fairness evaluation in breast MRI.
The MAMA-MIA challenge aims to advance the field with two complementary tasks: (1) Primary Tumour Segmentation and (2) Prediction of Pathologic Complete Response to Neoadjuvant Chemotherapy based on pretreatment dynamic contrast-enhanced MRI (DCE-MRI), leveraging imaging and clinical data. With 1,506 training cases from more than 20 clinical centres in the United States and a private set of over 570 test cases from three European centres (Spain, Poland and Lithuania), this challenge comprises the largest breast MRI dataset to date with 3D primary tumour segmentations approved by more than 16 experts from all over the world.
The challenge will incorporate metrics like Equal Opportunity Difference and Performance Disparities to measure biases across demographic subgroups, making it the first challenge to emphasize ethical AI deployment in breast cancer imaging.
Top 5 teams on each task and the winners of the Best Paper Award will be invited to contribute to a challenge summary publication.
The participation policies are desceribed in detail in the MICCAI Challenge Proposal.
Awards
Top 3 teams on each task will be invited to give oral presentations (in-person or online) during MICCAI 2025.
PRIMARY TUMOUR SEGMENTATION
1st Position
2nd Position
3rd Position
400 €
300 €
200 €
TrEATMENT RESPONSE PreDICTION
1st Position
2nd Position
3rd Position
400 €
300 €
200 €
BEST PAPER AWARD
This award celebrates innovative solutions that address generalizability and ensure equitable performance across demographic subgroups. To be eligible for the Best Paper Award, participants must submit their findings to the Deep-Breath Workshop. The best paper, as determined by the workshop and challenge committee, will receive:
Best Paper Award
300 €
Eleonora Poeta et. al. “Divergence-Aware Training with Automatic Subgroup Mitigation for Breast Tumor Segmentation“
CHALLENGE PREPRINT:
@misc{garrucho2026,
title={The MAMA-MIA Challenge: Advancing Generalizability and Fairness in Breast MRI Tumor Segmentation and Treatment Response Prediction},
author={Lidia Garrucho and Smriti Joshi and Kaisar Kushibar and Richard Osuala and Maciej Bobowicz and Xavier Bargalló and Paulius Jaruševičius and Kai Geissler and Raphael Schäfer and Muhammad Alberb and Tony Xu and Anne Martel and Daniel Sleiman and Navchetan Awasthi and Hadeel Awwad and Joan C. Vilanova and Robert Martí and Daan Schouten and Jeong Hoon Lee and Mirabela Rusu and Eleonora Poeta and Luisa Vargas and Eliana Pastor and Maria A. Zuluaga and Jessica Kächele and Dimitrios Bounias and Alexandra Ertl and Katarzyna Gwoździewicz and Maria-Laura Cosaka and Pasant M. Abo-Elhoda and Sara W. Tantawy and Shorouq S. Sakrana and Norhan O. Shawky-Abdelfatah and Amr Muhammad Abdo-Salem and Androniki Kozana and Eugen Divjak and Gordana Ivanac and Katerina Nikiforaki and Michail E. Klontzas and Rosa García-Dosdá and Meltem Gulsun-Akpinar and Oğuz Lafcı and Carlos Martín-Isla and Oliver Díaz and Laura Igual and Karim Lekadir},
year={2026},
eprint={2603.01250},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2603.01250},
}