PRIFLEX: A Secure Federated Learning Framework for Evaluating Privacy Leakage and Defense in Cross-Modal Medical Data
Article Sidebar
Main Article Content
Abstract
Federated learning (FL) enables decentralized, privacy-preserving machine learning by training models across distributed data without sharing raw patient information. However, most FL frameworks focus on unimodal data and overlook critical challenges in multimodal healthcare settings, such as privacy risks, fairness disparities, and reduced model interpretability. We present PRIFLEX (Privacy-Resilient Integration Framework for Learning Exchange), a novel FL framework designed for secure integration of structured and unstructured medical data. PRIFLEX combines 12-lead electrocardiograms (ECG) from the PhysioNet PTB-XL dataset and clinical notes from the Medical Information Mart for Intensive Care IV (MIMIC-IV), supporting early, late, and hybrid data fusion strategies. To safeguard model updates, it evaluates standalone and hybrid defenses using Differential Privacy (DP) and Secure Aggregation (SA) against gradient leakage, model inversion, and membership inference attacks. Results show that early fusion improves the area under the Receiver Operating Characteristic curve (AUROC) by up to 6.2%, while hybrid DP+SA reduces attack success rates by up to 84% and improves fairness with manageable system overhead. PRIFLEX also quantifies interpretability loss using SHapley Additive exPlanations (SHAP) and gradient-based methods, highlighting the trade-off between privacy and transparency. Overall, PRIFLEX sets a new benchmark for building secure, fair, and explainable federated learning systems in healthcare.
Article Details
Issue
Section
This work is licensed under a Creative Commons Attribution 4.0 International License.
How to Cite
References
[1] S. Saha, A. Hota, A. K. Chattopadhyay, A. Nag, and S. Nandi, “A multifaceted survey on privacy preservation of federated learning: progress, challenges, and opportunities,” Artif Intell Rev, vol. 57, no. 7, June 2024, doi: 10.1007/s10462-024-10766-7.
[2] S. Rajendran, W. Pan, M. R. Sabuncu, Y. Chen, J. Zhou, and F. Wang, “Learning across diverse biomedical data modalities and cohorts: Challenges and opportunities for innovation,” Patterns, vol. 5, no. 2, p. 100913, Feb. 2024, doi: 10.1016/j.patter.2023.100913.
[3] C. Celsa et al., “The application of artificial intelligence-based tools in the management of hepatocellular carcinoma: current status and future perspectives,” Hepatoma Res, Jan. 2025, doi: 10.20517/2394-5079.2024.126.
[4] B. McMahan, E. Moore, D. Ramage, S. Hampson, and B. Arcas, “Communication-Efficient Learning of Deep Networks from Decentralized Data,” in Proceedings of Machine Learning Research, 2017, pp. 1273–1282.
[5] M. J. Sheller et al., “Federated learning in medicine: facilitating multi-institutional collaborations without sharing patient data,” Sci Rep, vol. 10, no. 1, July 2020, doi: 10.1038/s41598-020-69250-1.
[6] G. A. Kaissis, M. R. Makowski, D. Rückert, and R. F. Braren, “Secure, privacy-preserving and federated machine learning in medical imaging,” Nat Mach Intell, vol. 2, no. 6, pp. 305–311, June 2020, doi: 10.1038/s42256-020-0186-1.
[7] N. I. Papandrianos, A. Feleki, S. Moustakidis, E. I. Papageorgiou, I. D. Apostolopoulos, and D. J. Apostolopoulos, “An Explainable Classification Method of SPECT Myocardial Perfusion Images in Nuclear Cardiology Using Deep Learning and Grad-CAM,” Applied Sciences, vol. 12, no. 15, p. 7592, July 2022, doi: 10.3390/app12157592.
[8] Y. Chen, X. Qin, J. Wang, C. Yu, and W. Gao, “FedHealth: A Federated Transfer Learning Framework for Wearable Healthcare,” IEEE Intell. Syst., vol. 35, no. 4, pp. 83–93, July 2020, doi: 10.1109/mis.2020.2988604.
[9] E.-S. M. El-Kenawy, M. M. Eid, H. L. Hussein, A. M. Osman, and A. M. Elshewey, “Optimized Deep Learning Model Using Binary Particle Swarm Optimization for Phishing Attack Detection: A Comparative Study,” Mesopotamian Journal of CyberSecurity, vol. 5, no. 2, pp. 685–703, July 2025, doi: https://doi.org/10.58496/MJCS/2025/041.
[10] L. Che, J. Wang, Y. Zhou, and F. Ma, “Multimodal Federated Learning: A Survey,” Sensors, vol. 23, no. 15, p. 6986, Aug. 2023, doi: 10.3390/s23156986.
[11] U. S. Begum, “Federated And Multi-Modal Learning Algorithms for Healthcare and Cross-Domain Analytics,” PIQM, vol. 1, no. 4, Nov. 2024, doi: 10.70023/sahd/241104.
[12] W. Huang, D. Wang, X. Ouyang, J. Wan, J. Liu, and T. Li, “Multimodal federated learning: Concept, methods, applications and future directions,” Information Fusion, vol. 112, p. 102576, Dec. 2024, doi: 10.1016/j.inffus.2024.102576.
[13] D. Li et al., “Blockchain for federated learning toward secure distributed machine learning systems: a systemic survey,” Soft Comput, vol. 26, no. 9, pp. 4423–4440, May 2022, doi: 10.1007/s00500-021-06496-5.
[14] M. Adam, A. Albaseer, U. Baroudi, and M. Abdallah, “Survey of Multimodal Federated Learning: Exploring Data Integration, Challenges, and Future Directions,” IEEE Open J. Commun. Soc., vol. 6, pp. 2510–2538, 2025, doi: 10.1109/ojcoms.2025.3554537.
[15] Y. Yan et al., “Cross-Modal Vertical Federated Learning for MRI Reconstruction,” IEEE J. Biomed. Health Inform., vol. 28, no. 11, pp. 6384–6394, Nov. 2024, doi: 10.1109/jbhi.2024.3360720.
[16] M. Fredrikson, S. Jha, and T. Ristenpart, “Model Inversion Attacks that Exploit Confidence Information and Basic Countermeasures,” in Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security, Denver Colorado USA: ACM, Oct. 2015, pp. 1322–1333. doi: 10.1145/2810103.2813677.
[17] N. Carlini et al., “Extracting Training Data from Large Language Models,” 2020, arXiv. doi: 10.48550/ARXIV.2012.07805.
[18] R. Madduri, Z. Li, T. Nandi, K. Kim, M. Ryu, and A. Rodriguez, “Advances in Privacy Preserving Federated Learning to Realize a Truly Learning Healthcare System,” in 2024 IEEE 6th International Conference on Trust, Privacy and Security in Intelligent Systems, and Applications (TPS-ISA), Washington, DC, USA: IEEE, Oct. 2024, pp. 273–279. doi: 10.1109/tps-isa62245.2024.00039.
[19] T.-H. Hwang, J. Shi, and K. Lee, “Enhancing Privacy-Preserving Personal Identification Through Federated Learning With Multimodal Vital Signs Data,” IEEE Access, vol. 11, pp. 121556–121566, 2023, doi: 10.1109/access.2023.3328641.
[20] N. Truong, K. Sun, S. Wang, F. Guitton, and Y. Guo, “Privacy preservation in federated learning: An insightful survey from the GDPR perspective,” Computers & Security, vol. 110, p. 102402, Nov. 2021, doi: 10.1016/j.cose.2021.102402.
[21] N. Rieke et al., “The future of digital health with federated learning,” npj Digit. Med., vol. 3, no. 1, Sept. 2020, doi: 10.1038/s41746-020-00323-1.
[22] M. Ghassemi, L. Oakden-Rayner, and A. L. Beam, “The false hope of current approaches to explainable artificial intelligence in health care,” The Lancet Digital Health, vol. 3, no. 11, pp. e745–e750, Nov. 2021, doi: 10.1016/s2589-7500(21)00208-9.
[23] E. Tjoa and C. Guan, “A Survey on Explainable Artificial Intelligence (XAI): Toward Medical XAI,” IEEE Trans. Neural Netw. Learning Syst., vol. 32, no. 11, pp. 4793–4813, Nov. 2021, doi: 10.1109/tnnls.2020.3027314.
[24] F. Zhang, Z. Shuai, K. Kuang, F. Wu, Y. Zhuang, and J. Xiao, “Unified fair federated learning for digital healthcare,” Patterns, vol. 5, no. 1, p. 100907, Jan. 2024, doi: 10.1016/j.patter.2023.100907.
[25] P. Barra, A. Della Greca, I. Amaro, A. Tortora, and M. Staffa, “A Comparative Analysis of XAI Techniques for Medical Imaging: Challenges and Opportunities,” in 2024 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Lisbon, Portugal: IEEE, Dec. 2024, pp. 6782–6788. doi: 10.1109/bibm62325.2024.10821983.
[26] Y.-M. Lin, Y. Gao, M.-G. Gong, S.-J. Zhang, Y.-Q. Zhang, and Z.-Y. Li, “Federated Learning on Multimodal Data: A Comprehensive Survey,” Mach. Intell. Res., vol. 20, no. 4, pp. 539–553, Aug. 2023, doi: 10.1007/s11633-022-1398-0.
[27] P. Dubey, P. Dubey, and P. N. Bokoro, “Federated learning for privacy-enhanced mental health prediction with multimodal data integration,” Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, vol. 13, no. 1, Dec. 2025, doi: 10.1080/21681163.2025.2509672.
[28] M. Nasr, R. Shokri, and A. Houmansadr, “Comprehensive Privacy Analysis of Deep Learning: Passive and Active White-box Inference Attacks against Centralized and Federated Learning,” in 2019 IEEE Symposium on Security and Privacy (SP), San Francisco, CA, USA: IEEE, May 2019, pp. 739–753. doi: 10.1109/sp.2019.00065.
[29] A. D. Salman and R. R. Al-Dahhan, “Ensure Privacy-Preserving Using Deep Learning,” Mesopotamian Journal of CyberSecurity, vol. 5, no. 2, pp. 703–720, July 2025, doi: https://doi.org/10.58496/MJCS/2025/042.
[30] C. Chen, Z. Zhou, P. Tang, L. He, and S. Su, “Enforcing group fairness in privacy-preserving Federated Learning,” Future Generation Computer Systems, vol. 160, pp. 890–900, Nov. 2024, doi: 10.1016/j.future.2024.06.040.
[31] T. U. Islam, R. Ghasemi, and N. Mohammed, “Privacy-Preserving Federated Learning Model for Healthcare Data,” in 2022 IEEE 12th Annual Computing and Communication Workshop and Conference (CCWC), Las Vegas, NV, USA: IEEE, Jan. 2022, pp. 0281–0287. doi: 10.1109/ccwc54503.2022.9720752.
[32] A. Holzinger, C. Biemann, C. S. Pattichis, and D. B. Kell, “What do we need to build explainable AI systems for the medical domain?,” Dec. 28, 2017, arXiv: arXiv:1712.09923. doi: 10.48550/arXiv.1712.09923.
[33] K. Wei et al., “Personalized Federated Learning With Differential Privacy and Convergence Guarantee,” IEEE Trans.Inform.Forensic Secur., vol. 18, pp. 4488–4503, 2023, doi: 10.1109/tifs.2023.3293417.
[34] X. Zhang, Y. Kang, K. Chen, L. Fan, and Q. Yang, “Trading Off Privacy, Utility, and Efficiency in Federated Learning,” ACM Trans. Intell. Syst. Technol., vol. 14, no. 6, pp. 1–32, Dec. 2023, doi: 10.1145/3595185.
[35] Dr. M. O. Nassar and F. F. Al-Mashagba, “Optimal Ensemble Learning with Meta-heuristics for Multiclass Classification of Syscall-Binder Interactions in Mobile Applications,” JoWUA, vol. 16, no. 1, pp. 26–48, Mar. 2025, doi: 10.58346/jowua.2025.i1.002.
[36] M. Alshinwan et al., “Dragonfly algorithm: a comprehensive survey of its results, variants, and applications,” Multimed Tools Appl, vol. 80, no. 10, pp. 14979–15016, Apr. 2021, doi: 10.1007/s11042-020-10255-3.
[37] F. Shannaq, M. Shehab, A. Alshorman, M. Hammad, B. Hammo, and W. Al‐Omari, “Exploring Metaheuristic Optimization Algorithms in the Context of Textual Cyberharassment: A Systematic Review,” Expert Systems, vol. 42, no. 2, Feb. 2025, doi: 10.1111/exsy.13826.
[38] M. Sh. Daoud, M. Shehab, L. Abualigah, and C.-L. Thanh, “Hybrid Modified Chimp Optimization Algorithm and Reinforcement Learning for Global Numeric Optimization,” J Bionic Eng, vol. 20, no. 6, pp. 2896–2915, Nov. 2023, doi: 10.1007/s42235-023-00394-2.
[39] M. Shehab and L. R. Alzabin, “Evaluating the Effectiveness of Stealth Protocols and Proxying in Hiding VPN Usage,” JCCE, Sept. 2024, doi: 10.47852/bonviewjcce42023642.
[40] T. Feng et al., “FedMultimodal: A Benchmark for Multimodal Federated Learning,” in Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, Long Beach CA USA: ACM, Aug. 2023, pp. 4035–4045. doi: 10.1145/3580305.3599825.
[41] Y. Jones, F. Deligianni, and J. Dalton, “Improving ECG Classification Interpretability using Saliency Maps,” in 2020 IEEE 20th International Conference on Bioinformatics and Bioengineering (BIBE), Cincinnati, OH, USA: IEEE, Oct. 2020, pp. 675–682. doi: 10.1109/bibe50027.2020.00114.
[42] J. Liu, Y. Li, M. Zhao, L. Liu, and N. Kumar, “EPFFL: Enhancing Privacy and Fairness in Federated Learning for Distributed E-Healthcare Data Sharing Services,” IEEE Trans. Dependable and Secure Comput., vol. 22, no. 2, pp. 1239–1252, Mar. 2025, doi: 10.1109/tdsc.2024.3431542.
[43] G. Wen and L. Li, “Federated transfer learning with differential privacy for multi-omics survival analysis,” Briefings in Bioinformatics, vol. 26, no. 2, Mar. 2025, doi: 10.1093/bib/bbaf166.
[44] D. K. Panda et al., “Creating intelligent cyberinfrastructure for democratizing AI,” AI Magazine, vol. 45, no. 1, pp. 22–28, Mar. 2024, doi: 10.1002/aaai.12166.
[45] D. Kim, H. Woo, and Y. Lee, “Addressing Bias and Fairness Using Fair Federated Learning: A Synthetic Review,” Electronics, vol. 13, no. 23, p. 4664, Nov. 2024, doi: 10.3390/electronics13234664.
[46] A. Bechar, R. Medjoudj, Y. Elmir, Y. Himeur, and A. Amira, “Federated and transfer learning for cancer detection based on image analysis,” Neural Comput & Applic, vol. 37, no. 4, pp. 2239–2284, Feb. 2025, doi: 10.1007/s00521-024-10956-y.
[47] K. Krishna Prakasha and U. Sumalatha, “Privacy-Preserving Techniques in Biometric Systems: Approaches and Challenges,” IEEE Access, vol. 13, pp. 32584–32616, 2025, doi: 10.1109/access.2025.3541649.
[48] C. Song, Z. Wang, W. Peng, and N. Yang, “Secure and Efficient Federated Learning Schemes for Healthcare Systems,” Electronics, vol. 13, no. 13, p. 2620, July 2024, doi: 10.3390/electronics13132620.
[49] N. Latif, W. Ma, and H. B. Ahmad, “Advancements in securing federated learning with IDS: a comprehensive review of neural networks and feature engineering techniques for malicious client detection,” Artif Intell Rev, vol. 58, no. 3, Jan. 2025, doi: 10.1007/s10462-024-11082-w.
[50] K. Khalil et al., “A Federated Learning Model Based on Hardware Acceleration for the Early Detection of Alzheimer’s Disease,” Sensors, vol. 23, no. 19, p. 8272, Oct. 2023, doi: 10.3390/s23198272.
[51] Y. Liu, J. Huang, Y. Li, D. Wang, and B. Xiao, “Generative AI model privacy: a survey,” Artif Intell Rev, vol. 58, no. 1, Dec. 2024, doi: 10.1007/s10462-024-11024-6.
[52] F. J. Piran, Z. Chen, M. Imani, and F. Imani, “Privacy-Preserving Federated Learning with Differentially Private Hyperdimensional Computing,” Computers and Electrical Engineering, vol. 123, p. 110261, Apr. 2025, doi: 10.1016/j.compeleceng.2025.110261.
[53] C. Anagnostopoulos, A. Gkillas, C. Mavrokefalidis, E.-V. Pikoulis, N. Piperigkos, and A. S. Lalos, “Multimodal Federated Learning in AIoT Systems: Existing Solutions, Applications, and Challenges,” IEEE Access, vol. 12, pp. 180864–180902, 2024, doi: 10.1109/access.2024.3508030.