Federated Deep Learning for Brain Tumor Segmentation: Comparative Analysis of U-Net Architectures

Abstract

Medical image segmentation is essential for precise diagnosis and treatment planning, particularly in brain tumor analysis where accurate boundary delineation directly impacts surgical outcomes and therapeutic decisions. However, privacy concerns and regulatory constraints limit data sharing across healthcare institutions, restricting the development of robust deep learning models that require large, diverse training datasets. We propose a federated deep learning framework that enables collaborative brain tumor segmentation model training without compromising patient data privacy. Our approach implements three U-Net architectural variants (standard U-Net, Residual U-Net, and Attention U-Net) across multi-institutional BraTS datasets (2019-2021) using the Federated Averaging algorithm. Performance evaluation on three tumor sub-regions (whole tumor, tumor core, enhancing tumor) demonstrates that federated Attention U-Net achieves Dice scores of 0.905, 0.859, and 0.811 respectively on BraTS 2021, representing only 0.8% degradation compared to centralized training while enabling privacy-preserving collaboration. Our federated approach preserves data privacy while achieving comparable accuracy to centralized methods, demonstrating comprehensive evaluation of attention-based federated learning for multi-institutional brain tumor segmentation with statistical significance validation across three consecutive challenge datasets.

Keywords

• Brain tumor segmentation
• Federated learning
• Attention mechanisms
• U-Net
• Medical imaging
• Multi-institutional learning

Supporting Institution

This research was supported by The Scientific and Technological Research Council of Türkiye (TÜBİTAK) under grant number 125E398 and constitutes part of the author’s master’s thesis work.

Project Number

TUBİTAK Grant number 125E398

Ethical Statement

This study was conducted using publicly available and fully anonymized BraTS challenge datasets (2019–2021). Since no private or identifiable patient information was used, additional ethical approval or informed consent was not required. All data were used strictly for academic research in accordance with the relevant guidelines and regulations.

Thanks

The authors thank the organizers of the BraTS challenges and all participating institutions for providing the datasets that made this study possible. We also acknowledge the neuroradiologists who contributed expert annotations and the MICCAI community for maintaining the benchmarking framework. Special gratitude is extended to the institutions and colleagues who supported the development of this research.

Beyin Tümörü Segmentasyonu için Federatif Derin Öğrenme: U-Net Mimarilerinin Karşılaştırmalı Analizi

Abstract

Keywords

• Beyin tümörü segmentasyonu
• Federatif öğrenme
• U-Net
• Tıbbi görüntüleme
• Çok kurumlu öğrenme

Supporting Institution

Bu araştırma, Türkiye Bilimsel ve Teknolojik Araştırma Kurumu (TÜBİTAK) tarafından 125E398 numaralı proje kapsamında desteklenmiştir ve yazarın yüksek lisans tezinin bir parçasını oluşturmaktadır.

Project Number

TUBİTAK Grant number 125E398

Ethical Statement

Bu çalışma, kamuya açık ve tamamen anonimleştirilmiş BraTS yarışma veri setleri (2019–2021) kullanılarak gerçekleştirilmiştir. Herhangi bir kişisel veya tanımlayıcı hasta verisi kullanılmadığından ek etik kurul iznine veya bilgilendirilmiş onama ihtiyaç duyulmamıştır. Tüm veriler, ilgili etik yönergeler ve düzenlemelere uygun şekilde yalnızca akademik araştırma amacıyla kullanılmıştır.

Thanks

Yazarlar, bu çalışmanın gerçekleştirilmesine olanak sağlayan BraTS organizatörlerine ve veri sağlayan tüm kurumlara teşekkür eder. Ayrıca uzman nöroradyologların sağladığı doğrulama anotasyonları ve MICCAI topluluğunun sağladığı değerlendirme altyapısı için teşekkür ederiz. Çalışmanın geliştirilme sürecinde destekleyen kurumlara ve tüm meslektaşlara özel olarak teşekkür ederiz.

References

1. [1] S. Bauer, R. Wiest, L. P. Nolte, and M. Reyes, "A survey of MRI-based medical image analysis for brain tumor studies," Phys. Med. Biol., vol. 58, no. 13, pp. R97–R129, 2013.W.-K. Chen, Linear Networks and Systems. Belmont, CA, USA: Wadsworth, 1993, pp. 123–135.
2. [2] B. H. Menze et al., "The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)," IEEE Trans. Med. Imaging, vol. 34, no. 10, pp. 1993–2024, 2015.
3. [3] O. Ronneberger, P. Fischer, and T. Brox, "U-Net: Convolutional networks for biomedical image segmentation," in Medical Image Computing and Computer-Assisted Intervention – MICCAI 2015, 2015, pp. 234–241.
4. [4] Z. Zhang, Q. Liu, and Y. Wang, "Road extraction by deep residual U-Net," IEEE Geosci. Remote Sens. Lett., vol. 15, no. 5, pp. 749–753, 2018.
5. [5] O. Oktay et al., "Attention U-Net: Learning where to look for the pancreas," arXiv preprint arXiv:1804.03999, 2018.
6. [6] S. Bakas et al., "Advancing the cancer genome atlas glioma MRI collections with expert segmentation labels and radiomic features," Sci. Data, vol. 4, pp. 170117, 2017.
7. [7] N. J. Tustison et al., "Optimal symmetric multimodal templates and concatenated random forests for supervised brain tumor segmentation (simplified) with ANTsR," Neuroinformatics, vol. 13, no. 2, pp. 209–225, 2015.
8. [8] F. Isensee et al., "nnU-Net: a self-adapting framework for U-Net-based medical image segmentation," Nat. Methods, vol. 18, pp. 203–211, 2021.

9. [9] A. Myronenko, "3D MRI brain tumor segmentation using autoencoder regularization," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, vol. 11383, 2019, pp. 311–320.
10. [10] European Union, "General Data Protection Regulation (GDPR)," Regulation (EU) 2016/679, 2016.
11. [11] U.S. Department of Health & Human Services, "Health Insurance Portability and Accountability Act (HIPAA)," 1996.
12. [12] S. Pereira, A. Pinto, V. Alves, and C. A. Silva, "Brain tumor segmentation using convolutional neural networks in MRI images," IEEE Trans. Med. Imaging, vol. 35, no. 5, pp. 1240–1251, 2016.
13. [13] B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. A. y. Arcas, "Communication-efficient learning of deep networks from decentralized data," in Proc. 20th Int. Conf. Artificial Intelligence and Statistics (AISTATS), 2017, pp. 1273–1282.
14. [14] R. C. Geyer, T. Klein, and M. Nabi, "Differentially private federated learning: A client level perspective," arXiv preprint arXiv:1712.07557, 2017.
15. [15] P. Kairouz et al., "Advances and open problems in federated learning," Found. Trends Mach. Learn., vol. 14, no. 1–2, pp. 1–210, 2021.
16. [16] M. J. Sheller et al., "Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data," Sci. Rep., vol. 10, pp. 12598, 2020.
17. [17] S. Silva et al., "Federated learning in breast density classification," arXiv preprint arXiv:1910.02135, 2019.
18. [18] I. Dayan et al., "Federated learning for predicting clinical outcomes in patients with COVID-19," Nat. Med., vol. 27, pp. 1735–1743, 2021.
19. [19] J. Xu et al., "Federated learning for ophthalmology: a new collaborative paradigm," Ophthalmol. Sci., vol. 1, no. 1, pp. 100017, 2021.
20. [20] M. J. Sheller, G. A. Reina, B. Edwards, J. Martin, and S. Bakas, "Multi-institutional deep learning modeling without sharing patient data: A feasibility study on brain tumor segmentation," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, Springer, Cham, 2019, pp. 92–104.
21. [21] W. Li et al., "Privacy-preserving federated brain tumour segmentation," in Machine Learning in Medical Imaging, Springer, Cham, 2019, pp. 133–141.
22. [22] P. Guo, P. Wang, J. Zhou, S. Jiang, and V. M. Patel, "Multi-institutional collaborations for improving deep learning-based magnetic resonance image reconstruction using federated learning," in Proc. IEEE/CVF Conf. Computer Vision and Pattern Recognition, 2021, pp. 2423–2432.
23. [23] R. Kumar, S. Maity, S. Saha, and S. K. Bandyopadhyay, "Brain tumor segmentation using blockchain-federated learning and 3D U-Net," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, Springer, Cham, 2020, pp. 450–462.
24. [24] L. Yi, J. Zhang, R. Zhang, J. Shi, G. Wang, and X. Liu, "SU-Net: An efficient encoder-decoder model of federated learning for brain tumor segmentation," in Artificial Neural Networks and Machine Learning – ICANN 2020, Springer, Cham, 2020, pp. 761–773.
25. [25] C. Liu et al., "FedMEPD: Federated multi-modal encoders with partially personalized decoders for brain tumor segmentation," Med. Image Anal., vol. 97, pp. 103274, 2025.
26. [26] M. Manthe, S. Duffner, and C. Lartizien, "Federated brain tumor segmentation: An extensive benchmark," Med. Image Anal., vol. 97, pp. 10 A. Asmita and P. Mittal, "Enhancing traditional learning mechanism of brain tumor segmentation for BraTS 23 employing contrastive learning," Biomed. Signal Process. Control, vol. 112, pp. 108460, 2025.3270, 2024.
27. [27] A. Asmita and P. Mittal, "Enhancing traditional learning mechanism of brain tumor segmentation for BraTS 23 employing contrastive learning," Biomed. Signal Process. Control, vol. 112, pp. 108460, 2025.
28. [28] Z. Zhou et al., "Comprehensive exploiting local and global features for brain tumor segmentation: A gated dual-branch hybrid attention mechanism," Biomed. Signal Process. Control, vol. 110, pp. 108285, 2025.
29. [29] W. Liu, J. Hou, X. Zhong, H. Zhan, J. Cheng, and G. Yue, "Improving multi-modal brain tumor segmentation via pre-training and knowledge distillation based post-training," Neurocomputing, vol. 640, pp. 130318, 2025.
30. [30] J. Su, Z. Luo, C. Wang, S. Lian, X. Lin, and S. Li, "Reconstruct incomplete relation for incomplete modality brain tumor segmentation," Neural Networks, vol. 180, pp. 106657, 2024.
31. [31] R. Lei, Y. Zhu, H. Liu, Y. Xu, and Q. Cai, "FE-Unet: A lightweight brain tumor segmentation network combining frequency domain processing and edge preservation," Biomed. Signal Process. Control, vol. 110, pp. 108274, 2025.
32. [32] M. F. Ahamed et al., "Classification and segmentation on multi-regional brain tumors using volumetric images of MRI with customized 3D U-Net framework," Comput. Med. Imaging Graph., vol. 110, pp. 102313, 2023.
33. [33] J. Nalepa, M. Marcinkiewicz, and M. Kawulok, "Data augmentation for brain-tumor segmentation: a review," Front. Comput. Neurosci., vol. 13, pp. 83, 2019.
34. [34] A. Chowdhury, H. Kassem, N. Padoy, and A. Karargyris, "A review of medical federated learning: Applications in oncology and cancer research," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, Springer, Cham, 2022, pp. 3–24.
35. [35] R. Ranjbarzadeh, A. Caputo, E. B. Tirkolaee, and S. Jafarzadeh Ghoushchi, "Brain tumor segmentation of MRI images: A comprehensive review on the application of artificial intelligence tools," Comput. Methods Programs Biomed., vol. 152, pp. 106405, 2023.
36. [36] A. Wadhwa, A. Bhardwaj, and V. S. Verma, "A review on brain tumor segmentation of MRI images," Magn. Reson. Imaging, vol. 61, pp. 247–259, 2019.
37. [37] S. Nalawade et al., "Federated learning for brain tumor segmentation using MRI and transformers," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, 2021, pp. 444–456.
38. [38] A. Tuladhar et al., "Federated learning using variable local training for brain tumor segmentation," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, 2022, pp. 298–309.
39. [39] M. I. Khan et al., "Adaptive weight aggregation in federated learning for brain tumor segmentation," in Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries, 2022, pp. 156–167.
40. [40] X. Xu et al., "Federated cross learning for medical image segmentation," IEEE Trans. Med. Imaging, vol. 41, no. 6, pp. 1572–1583, 2022.
41. [41] D. H. Mahlool and M. H. Abed, "Survey on federated learning: Concept and applications," in Mobile Computing and Sustainable Informatics, 2022, pp. 337–351.
42. [42] W. Dai et al., "LACCA: A lightweight adaptive clustered client aggregation for federated learning," IEEE Trans. Netw. Sci. Eng., vol. 11, no. 2, pp. 1945–1958, 2024.