The Federated Tumor Segmentation (FeTS) Challenge 2024: Efficient and Robust Aggregation Methods

Akis Linardos1,2Orcid, Sarthak Pati1,2,3, Ujjwal Baid1,2Orcid, Brandon Edwards4, Patrick Foley4, Kevin Ta4, Verena Chung5, Micah Sheller3,4, Muhammad Irfan Khan6, Mojtaba Jafaritadi7, Elina Kontio6, Suleiman Khan6, Leon Mächler8, Ivan Ezhov9, Suprosanna Shit9, Johannes C. Paetzold10, Gustav Grimberg11, Manuel A. Nickel9, David Naccache8, Vasilis Siomos12, Jonathan Passerat-Palmbach13, Giacomo Tarroni14,13, Daewoon Kim15, Leonard L. Klausmann16, Prashant Shah4, Bjoern Menze17, Dimitrios Makris18Orcid, Spyridon Bakas1,2,3,18,19,20Orcid
1: Division of Computational Pathology, Department of Pathology and Laboratory Medicine, Indiana University School of Medicine, Indianapolis, IN, USA, 2: Center for Federated Learning in Medicine, Indiana University School of Medicine, Indianapolis, IN, USA, 3: Medical AI Group, MLCommons, San Francisco, CA, USA, 4: Intel Corporation, Santa Clara, CA, USA, 5: Sage Bionetworks, Seattle, WA, USA, 6: Turku University of Applied Sciences, Turku, Finland, 7: Stanford University, Stanford, CA, USA, 8: Ecole Normale Supérieure, Paris, France, 9: Technical University of Munich, Munich, Germany, 10: Weill Cornell Medicine, New York, USA, 11: Ezri AI Labs, Paris, France, 12: City St George’s, University of London, UK, 13: Imperial College London, London, UK, 14: St George’s, University of London, UK, 15: Seoul National University, Seoul, South Korea, 16: Ostbayerische Technische Hochschule (OTH) Regensburg, Germany, 17: Universität Zürich, Zürich, Switzerland, 18: Kingston University London, London, UK, 19: Departments of Radiology and Imaging Sciences; Neurological Surgery; Biostatistics and Health Data Science, Indiana University School of Medicine, Indianapolis, IN, USA, 20: Department of Computer Science, Luddy School of Informatics, Computing and Engineering, Indiana University, Indianapolis, IN, USA
Publication date: 2025/12/05
https://doi.org/10.59275/j.melba.2025-5242
PDF · Code

Abstract

We present the design and results of the MICCAI Federated Tumor Segmentation (FeTS) Challenge 2024, which focuses on federated learning (FL) for glioma sub-region segmentation in multi-parametric MRI scans. Unlike previous FeTS challenges, this iteration exclusively evaluates novel weight aggregation methods for increased robustness and efficiency. Participating methods from six teams are evaluated using a standardized FL setup and a multi-institutional dataset derived from the BraTS glioma benchmark—a dataset consisting of 1,251 training cases, 219 validation cases, and 570 hidden test cases, with segmentations of enhancing tumor (ET), tumor core (TC), and whole tumor (WT). Teams are ranked by a cumulative scoring system that accounts for segmentation performance—measured by Dice Similarity Coefficient (DSC) and 95th percentile Hausdorff Distance (HD95)—and communication efficiency assessed through the convergence score. A PID-controller-based approach emerges as the top-performing method, achieving a mean DSC of 0.733, 0.761, and 0.751 for ET, TC, and WT, respectively, with corresponding HD95 values of 33.922mm, 33.623mm, and 32.309mm, while also being the most efficient with a convergence score of 0.764. These results contribute to ongoing advances in FL, building on top-performers from previous iterations of the challenge and surpassing them, highlighting the potential of PID controllers as a powerful mechanism for stabilizing and optimizing weight aggregation in FL. The challenge code is available at https://github.com/FeTS-AI/Challenge

Keywords

federated learning · biomedical challenge · segmentation · aggregation · brain tumor · glioma · glioblastoma

Bibtex @article{melba:2025:033:linardos, title = "The Federated Tumor Segmentation (FeTS) Challenge 2024: Efficient and Robust Aggregation Methods", author = "Linardos, Akis and Pati, Sarthak and Baid, Ujjwal and Edwards, Brandon and Foley, Patrick and Ta, Kevin and Chung, Verena and Sheller, Micah and Khan, Muhammad Irfan and Jafaritadi, Mojtaba and Kontio, Elina and Khan, Suleiman and Mächler, Leon and Ezhov, Ivan and Shit, Suprosanna and Paetzold, Johannes C. and Grimberg, Gustav and Nickel, Manuel A. and Naccache, David and Siomos, Vasilis and Passerat-Palmbach, Jonathan and Tarroni, Giacomo and Kim, Daewoon and Klausmann, Leonard L. and Shah, Prashant and Menze, Bjoern and Makris, Dimitrios and Bakas, Spyridon", journal = "Machine Learning for Biomedical Imaging", volume = "3", issue = "December 2025 issue", year = "2025", pages = "757--774", issn = "2766-905X", doi = "https://doi.org/10.59275/j.melba.2025-5242", url = "https://melba-journal.org/2025:033" }
RISTY - JOUR AU - Linardos, Akis AU - Pati, Sarthak AU - Baid, Ujjwal AU - Edwards, Brandon AU - Foley, Patrick AU - Ta, Kevin AU - Chung, Verena AU - Sheller, Micah AU - Khan, Muhammad Irfan AU - Jafaritadi, Mojtaba AU - Kontio, Elina AU - Khan, Suleiman AU - Mächler, Leon AU - Ezhov, Ivan AU - Shit, Suprosanna AU - Paetzold, Johannes C. AU - Grimberg, Gustav AU - Nickel, Manuel A. AU - Naccache, David AU - Siomos, Vasilis AU - Passerat-Palmbach, Jonathan AU - Tarroni, Giacomo AU - Kim, Daewoon AU - Klausmann, Leonard L. AU - Shah, Prashant AU - Menze, Bjoern AU - Makris, Dimitrios AU - Bakas, Spyridon PY - 2025 TI - The Federated Tumor Segmentation (FeTS) Challenge 2024: Efficient and Robust Aggregation Methods T2 - Machine Learning for Biomedical Imaging VL - 3 IS - December 2025 issue SP - 757 EP - 774 SN - 2766-905X DO - https://doi.org/10.59275/j.melba.2025-5242 UR - https://melba-journal.org/2025:033 ER -
2025:033 cover