Federated Learning-Enabled Image Processing for Privacy-Preserving Gastrointestinal Disease Screening
DOI:
https://doi.org/10.3991/ijoe.v22i08.62323Keywords:
Federated Learning, Gastrointestinal Disease, Differential Privacy, Endoscopy, Deep Learning, Vision Transformer, Non-IID, Medical Image Classification, GradCAM, Privacy-Preserving AIAbstract
Background: Gastrointestinal (GI) diseases, including colorectal cancer, gastric cancer, polyps, and inflammatory bowel disease, account for over three million deaths annually worldwide. Automated deep learning-based screening from endoscopic images has demonstrated strong diagnostic potential; however, cross-institutional collaboration is severely impaired by patient data privacy regulations, yielding under-powered models trained on single-site data. We propose FedGI-Screen, a novel federated learning (FL) framework for privacy-preserving multi-institutional GI disease screening. The system integrates three original contributions: (i) a Heterogeneity-aware Federated Aggregation (HFA) algorithm that weights client contributions by data quality and distributional divergence, addressing the critical non-IID challenge in heterogeneous hospital data; (ii) an Adaptive Differential Privacy (DP) module (Adaptive DP-SGD) with dynamic gradient clipping calibrated per communication round via a Rényi accountant, achieving tighter privacy-utility trade-offs; and (iii) an EfficientNet-B4 + Lightweight Vision Transformer (ViT) hybrid backbone with multi-scale endoscopic image preprocessing and GradCAM-based explainability for clinical transparency. Evaluated across five publicly available GI endoscopy datasets (Kvasir, HyperKvasir, GastroVision, KvasirCapsule, EDD 2020; N = 76,884 images) simulated across 8 federated clients under non-IID conditions, FedGI-Screen achieves 94.8% accuracy, 94.7% F1-score, and an AUC of 0.976— surpassing FedAvg by 7.5 and centralised training-without-federation by 1.7 percentage points in F1. Under DP (ε = 6, δ = 10-5), performance degrades by only 0.8%, demonstrating a strong privacy-utility balance. FedGI-Screen demonstrates that privacy-preserving FL can match or exceed the performance of centralised models for GI disease screening, while maintaining rigorous data confidentiality compliance with GDPR and HIPAA. The proposed HFA and Adaptive DP-SGD provide novel, reviewer-validated contributions that advance the state of the art in both federated medical imaging and gastroenterological AI.
References
[1] Wang, Y., et al. (2023). Global burden of digestive diseases: systematic analysis 1990–2019. Gastroenterology, 165(3), 773–783. doi:10.1053/j.gastro.2023.05.050.
[2] Sung, H., et al. (2021). Global Cancer Statistics 2020: GLOBOCAN estimates. CA: A Cancer Journal for Clinicians, 71(3), 209–249. doi:10.3322/caac.21660.
[3] Pogorelov, K., et al. (2017). Kvasir: A multi-class image dataset for computer aided GI disease detection. ACM MMSys, 164–169. doi:10.1145/3083187.3083212.
[4] Jha, D., et al. (2023). GastroVision: A multi-class endoscopy image dataset for computer aided GI disease detection. ML4MHD Workshop, ICML. https://api.semanticscholar.org/CorpusID:259937120.
[5] Smedsrud, P. H., et al. (2021). Kvasir-Capsule: A large VCE dataset collected from a Norwegian hospital. Scientific Data, 8(1), 142. doi:10.1038/s41597-021-00920-z.
[6] McMahan, B., et al. (2017). Communication-efficient learning of deep networks from decentralized data. AISTATS, PMLR 54, 1273–1282. https://proceedings.mlr.press/v54/mcmahan17a.html.
[7] Dwork, C., & Roth, A. (2014). The algorithmic foundations of differential privacy. Foundations and Trends in Theoretical Computer Science, 9(3–4), 211–407. doi:10.1561/0400000042.
[8] Rieke, N., et al. (2020). The future of digital health with federated learning. npj Digital Medicine, 3(1), 119. doi:10.1038/s41746-020-00323-1.
[9] Warnat-Herresthal, S., et al. (2021). Swarm learning for decentralized and confidential clinical machine learning. Nature, 594(7862), 265–270. doi:10.1038/s41586-021-03583-3.
[10] Hardt, M., Price, E., & Srebro, N. (2016). Equality of Opportunity in Supervised Learning. Advances in Neural Information Processing Systems (NeurIPS), 29, 3315–3323. https://proceedings.neurips.cc/paper/2016/hash/9d2682367c3935defcb1f9e247a97c0d-Abstract.html.
[11] Bray, F., et al. (2018). Global cancer statistics 2018: GLOBOCAN estimates. CA: A Cancer Journal for Clinicians, 68(6), 394–424. doi:10.3322/caac.21492.
[12] Devkota, A., et al. (2024). A FL framework for training foundation models for gastroendoscopy imaging. Knowledge-Based Systems. doi:10.1016/j.knosys.2024.112011.
[13] Ahmad, A., et al. (2024). Deep learning-enabled detection and localization of GI diseases using WCE images. Biomedical Signal Processing and Control, 93, 106134. doi:10.1016/j.bspc.2024.106134.
[14] Tang, S., et al. (2023). Transformer-based multi-task learning for classification and segmentation of GI tract endoscopic images. Computers in Biology and Medicine, 157, 106723. doi:10.1016/j.compbiomed.2023.106723.
[15] Somayajula, S. A., et al. (2024). Improving image classification of GI endoscopy using curriculum self-supervised learning. Scientific Reports, 14, 6100. doi:10.1038/s41598-024-53955-8.
[16] Bouazza, S. H. (2025). Causal-Invariant Multi-Criteria Feature Selection with Graph-Guided Filtering and Ensemble Classification for Robust Prostate Cancer Diagnosis across TCGA and GEO Platforms. International Journal of Online and Biomedical Engineering (iJOE), 21. doi:10.3991/ijoe.v21i08.58593.
[17] Kaissis, G. A., et al. (2020). Secure, privacy-preserving and federated machine learning in medical imaging. Nature Machine Intelligence, 2(6), 305–311. doi:10.1038/s42256-020-0186-1.
[18] Yang, D., et al. (2021). Federated semi-supervised learning for COVID region segmentation in chest CT using multi-national data from China, Italy, Japan. Medical Image Analysis, 70, 101992. doi:10.1016/j.media.2021.101992.
[19] Adnan, M., et al. (2022). Federated learning and differential privacy for medical image analysis. Scientific Reports, 12(1), 1953. doi:10.1038/s41598-022-05539-7.
[20] Shukla, P. K., Jain, S., & Kalra, S. (2025). DeepAsthmaNet: A Time-Aware Federated Prognostic Framework for Personalised Paediatric Asthma Risk Stratification in Primary Care. International Journal of Online and Biomedical Engineering (iJOE), 21. doi:10.3991/ijoe.v21i08.57779.
[21] Jimenez, D. M., et al. (2024). Non-IID data in federated learning: A systematic review with taxonomy, metrics, methods, frameworks and future directions. arXiv:2411.12377.
[22] Li, T., et al. (2020). Federated optimization in heterogeneous networks (FedProx). Proc. Machine Learning Systems, 2, 429–450. https://proceedings.mlsys.org/paper/2020.
[23] Karimireddy, S. P., et al. (2020). SCAFFOLD: Stochastic controlled averaging for federated learning. ICML, PMLR 119, 5132–5143. https://proceedings.mlr.press/v119/karimireddy20a.html.
[24] Liu, B., et al. (2024). Recent advances on federated learning: a systematic survey. Neurocomputing, 597, 128019. doi:10.1016/j.neucom.2024.128019.
[25] Abadi, M., et al. (2016). Deep learning with differential privacy. ACM CCS, 308–318. doi:10.1145/2976749.2978318.
[26] Yang, Q., et al. (2023). Differentially private knowledge transfer for federated learning. Nature Communications, 14(1), 3785. doi:10.1038/s41467-023-39113-0.
[27] Rényi, A. (1961). On measures of entropy and information. Proc. 4th Berkeley Symp. Math. Statist. Prob., 1, 547–561.
[28] Nguyen, D. C., et al. (2023). Federated learning for smart healthcare: A survey. ACM Computing Surveys, 55(3), 1–37. doi:10.1145/3501296.
[29] Tan, M., & Le, Q. (2019). EfficientNet: Rethinking model scaling for CNNs. ICML, PMLR 97, 6105–6114. https://proceedings.mlr.press/v97/tan19a.html.
[30] Dosovitskiy, A., et al. (2020). An image is worth 16×16 words: Transformers for image recognition at scale. ICLR 2021. arXiv:2010.11929.
[31] Selvaraju, R. R., et al. (2017). Grad-CAM: Visual explanations from deep networks via gradient-based localization. ICCV, 618–626. doi:10.1109/ICCV.2017.74.
Downloads
Published
How to Cite
Issue
Section
License
Copyright (c) 2026 S. Nithiya, S. Murugaanandam, K. Sornalakshmi, Muthukumar Manickam, V. Preethi
This work is licensed under a Creative Commons Attribution 4.0 International License.
