Ransomware attacks—especially during the COVID-19 pandemic—have placed healthcare systems under immense strain, accounting for up to 85% of all malware incidents. While several ML-based detection methods exist, they largely rely on labelled datasets and known ransomware signatures, making them ill-suited to the dynamic, evolving nature of threats in Internet of Health Things (IoHTs) ecosystems.
Fed Health developed a federated learning (FL) framework that enables ransomware detection while preserving data privacy across multiple healthcare networks. The core innovation lies in achieving robust ransomware detection with minimal communication overhead by integrating few-shot learning and variance reduction techniques.
The key objectives of Fed Health are:
1.Fed Health allows for fast, adaptive, and privacy-preserving ransomware detection across distributed clinical infrastructures without compromising accuracy.
2. Framework trained on containing ransomware in healthcare environments.
3.Fed Health enables Ransomware detection while preserving data privacy across multiple healthcare networks.
Expected Impact
1.Enhances Cybersecurity in Healthcare: Enables timely, privacy-preserving detection of both known and novel ransomware attacks in distributed IoHTs.
2.Improves Federated Learning Efficiency: Reduces communication rounds while maintaining performance, addressing key FL deployment bottlenecks.
3.Supports Scalable Real-World Adoption: Produces a robust, industry-ready solution that integrates seamlessly into existing healthcare IT systems, enabling rapid uptake across hospital networks and critical infrastructure.
