Privacy-Preserving Federated Learning Model for Healthcare Cybersecurity

Abstract

The rapid digital transformation of healthcare systems has significantly improved patient care and operational efficiency. However, it has also increased the risk of cyberattacks, data breaches, and unauthorized access to sensitive medical information. Traditional centralized machine learning approaches require the collection of data from multiple healthcare institutions, creating privacy concerns and regulatory compliance challenges. To address these issues, this study proposes a Privacy-Preserving Federated Learning Model for Healthcare Cybersecurity that enables collaborative model training without sharing raw patient data. The proposed framework utilizes federated learning to distribute model training across healthcare organizations while incorporating privacypreserving mechanisms such as secure aggregation and differential privacy to protect sensitive information. The model is designed to detect and mitigate cybersecurity threats, including intrusion attempts, malware attacks, and anomalous network activities in healthcare environments. Experimental evaluation demonstrates that the proposed approach achieves high detection accuracy while maintaining data confidentiality and reducing privacy risks. Comparative analysis indicates that the federated learning-based framework provides improved security, scalability, and compliance with healthcare data protection regulations compared to conventional centralized methods. The results highlight the potential of privacy-preserving federated learning as an effective solution for enhancing cybersecurity resilience in modern healthcare systems.