Interpreting Large-Scale Attacks Against Open-Source Medical Systems Using eXplainable AI

The recent increase in IoMT has rapidly changed the healthcare industry. Its use in hospitals, however, has also raised severe security and privacy concerns. Integrating machine learning algorithms for predicting and identifying potential cyber threats represents a promising advancement. They, however, were not widely accepted in medical practice because of their inherent complexity and lack of explainability. These constraints make implementing robust security systems challenging. In this paper, we propose an explainable artificial intelligence-based intrusion detection system, including an XGB-based detector and a SHAP-based explainer, to enhance the security of IoMT devices. The study is significant because it comprehensively addresses the assessment of large-scale attacks against medical devices with interpretation and explanation. Its outcomes improve healthcare delivery, reduce treatment errors, and improve patient trust.