Work

Federated Machine Learning for Collaborative Cybersecurity in Distributed E-Health Systems: A Privacy-Preserving Approach for Healthcare Network Security

Year: 2025
Type: article
Abstract: The proliferation of distributed electronic health systems has created unprecedented opportunities for healthcare delivery while simultaneously introducing complex cybersecurity vulnerabilities that traditional centralized security approaches cannot adequately address. This research presents a novel federated machine learning framework specifically designed for collaborative cybersecurity in distributed e-health environments, enabling multiple healthcare institutions to jointly develop robust threat detection capabilities without compromising patient data privacy or regulatory compliance. Our proposed system leverages federated learning principles to train machine learning models across distributed healthcare nodes, where each participating institution contributes to the collective security intelligence while maintaining strict data locality and privacy constraints. The experimental evaluation demonstrates that our federated intrusion detection system achieves 93.7% accuracy in identifying anomalous network behavior while maintaining full HIPAA compliance and reducing false positive rates by 34% compared to traditional centralized approaches. The framework successfully addresses critical challenges including model poisoning attacks, communication overhead optimization, and non-independent and identically distributed data distributions across healthcare providers through innovative secure aggregation protocols and differential privacy mechanisms. Implementation results across a simulated network of 15 healthcare institutions show significant improvements in detecting sophisticated threats including ransomware, advanced persistent threats, and zero-day exploits while preserving institutional data sovereignty. This research establishes a foundation for privacy-preserving collaborative cybersecurity frameworks that enable healthcare institutions to collectively enhance their security posture without compromising patient confidentiality or violating regulatory requirements.
Language: English
Open Access status: diamond
OpenAlex ID: W7216714378
FWCI:
Cites: 27
Cited by:
Related to: 10