Logo PTI Logo FedCSIS

Proceedings of the 21st Conference on Computer Science and Intelligence Systems (FedCSIS)

Annals of Computer Science and Information Systems, Volume 47

Federated Learning for Histopathological Image Segmentation: A Privacy-Preserving Study Using the EBHI-Seg Dataset

,

DOI: http://dx.doi.org/10.15439/2026F7491

Citation: Mohammadreza Azimi,

Full text

Abstract. The increasing adoption of artificial intelligence in digital pathology is often constrained by limited data sharing due to privacy and regulatory concerns. Federated learning (FL) has emerged as a promising paradigm that enables collaborative model training across institutions without centralizing sensitive data. In this study, we investigate the feasibility and performance of federated learning for histopathological image segmentation using the EBHI-Seg dataset. We simulate a multi-institutional setting by partitioning the dataset into multiple decentralized clients, reflecting realistic clinical data distributions under both independent and non-identically distributed (non-IID) conditions. A convolutional neural network based on the U-Net architecture is employed as the baseline segmentation model and trained using a federated optimization strategy. Model performance is evaluated using standard metrics, including Dice similarity coefficient and Intersection over Union (IoU), and compared against a centrally trained counterpart. Our results demonstrate that federated learning achieves competitive segmentation performance while preserving data locality, although performance degradation is observed under highly heterogeneous data distributions. Further analysis highlights the impact of client variability, data imbalance, and domain shifts on model convergence and generalization. These findings underscore the potential of federated learning as a viable framework for privacy-preserving collaborative medical image analysis and provide insights into its practical deployment in digital pathology.

References

  1. Y. Tian, Y. Zheng, Y. Liu, and Y. Wang, “GDC-Net: a U-Net for precise brain vessel segmentation with global–local and depthwise attention plus content-aware upsampling,” The Journal of Supercomputing, vol. 82, article 215, 2026, http://dx.doi.org/10.1007/s11227-026-08376-x.
  2. S. M. Turjya and M. Fawakherji, “Federated lung nodule segmentation using a hybrid transformer–U-Net architecture,” Scientific Reports, vol. 16, article 5228, 2026, http://dx.doi.org/10.1038/ s41598-026-35243-9.
  3. C. Chauhan et al., “The critical role of standards for AI in digital pathology: Digital Pathology Association Concept Paper,” Journal of Pathology Informatics, vol. 21, article 100645, 2026, http://dx.doi.org/10.1016/j.jpi.2026.100645.
  4. B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. Agüera y Arcas, “Communication-efficient learning of deep networks from decentralized data,” in Proc. 20th International Conference on Artificial Intelligence and Statistics (AISTATS), PMLR, vol. 54, pp. 1273–1282, 2017. Available: https://proceedings.mlr.press/v54/mcmahan17a.html.
  5. S. Bakas, X. Li, P. Shah, and H. R. Roth, “Federated learning in healthcare: From research to real-world deployment,” Annual Review of Biomedical Engineering, vol. 28, 2026, http://dx.doi.org/10.1146/ annurev-bioeng-080125-041414.
  6. E. Markodimitrakis et al., “Federated learning on magnetic resonance imaging: a critical review,” Artificial Intelligence Review, vol. 59, article 97, 2026, http://dx.doi.org/10.1007/s10462-026-11508-7.
  7. A. R. Aurnob, S. A. Tanim, T. E. Shrestha, M. F. Mridha, and D. Mistry, “FedFusionNet: Advancing oral cancer recurrence prediction through federated fusion modeling,” Information Fusion, vol. 132, article 104205, 2026, http://dx.doi.org/10.1016/j.inffus.2026.104205.
  8. H. Guan, P. T. Yap, A. Bozoki, and M. Liu, “Federated learning for medical image analysis: A survey,” Pattern Recognition, vol. 151, article 110424, 2024, http://dx.doi.org/10.1016/j.patcog.2024.110424.
  9. S. Nazir and M. Kaleem, “Federated learning for medical image analysis with deep neural networks,” Diagnostics, vol. 13, no. 9, article 1532, 2023, http://dx.doi.org/10.3390/diagnostics13091532.
  10. M. H. U. Rehman, W. H. L. Pinaya, P. Nachev, J. T. Teo, S. Ourselin, and M. J. Cardoso, “Federated learning for medical imaging radiology,” The British Journal of Radiology, vol. 96, no. 1150, article 20220890, 2023, http://dx.doi.org/10.1259/bjr.20220890.
  11. D. Ghosh, M. Mehjabin, M. E. Rayed, M. F. Mridha, and M. M. Kabir, “Advancements and challenges of federated learning in medical imaging: a systematic literature review,” Artificial Intelligence Review, vol. 59, article 87, 2026, http://dx.doi.org/10.1007/s10462-025-11489-z.
  12. V. S. Parekh, S. Lai, V. Braverman, J. Leal, S. Rowe, J. J. Pillai, and M. A. Jacobs, “Cross-domain federated learning in medical imaging,” arXiv preprint https://arxiv.org/abs/2112.10001, 2021, http://dx.doi.org/10.48550/ arXiv.2112.10001.
  13. S. Alphonse, F. Mathew, K. Dhanush, and V. Dinesh, “Federated learning with integrated attention multiscale model for brain tumor segmentation,” Scientific Reports, vol. 15, article 11889, 2025, http://dx.doi.org/10.1038/s41598-025-96416-6.
  14. M. Sun, Z. Yang, Y. Huang, H. Yu, Y. Chen, S. Qi, A. B. J. Teoh, and Y. Zhang, “Federated learning for large models in medical imaging: a comprehensive review,” arXiv preprint arXiv:2508.20414, 2025, http://dx.doi.org/10.48550/arXiv.2508.20414.
  15. L. Shi et al., “EBHI-Seg: a novel enteroscope biopsy histopathological hematoxylin and eosin image dataset for image segmentation tasks,” Frontiers in Medicine, vol. 10, article 1114673, 2023, http://dx.doi.org/10.3389/fmed.2023.1114673.