PULSE: Leakage Aware, Physiology Informed and Conformally Risk Controlled Intrusion Detection for the Internet of Medical Things

PULSE is a research project focused on improving intrusion detection and security in Internet of Medical Things (IoMT) environments. The study presents a leakage-aware and episode-aware evaluation framework and proposes PULSE, a detection approach combining network traffic analysis, physiological consistency monitoring, and conformal risk control. The research evaluates the approach using publicly available WUSTL EHMS 2020 and ToN IoT datasets. It examines intrusion detection performance, attack episode detection, uncertainty handling, temporal drift, and physiological data integrity. This record contains the research manuscript and the accompanying code implementation, including scripts, configurations, and result files intended to support reproducibility and further research. Keywords: Internet of Medical Things (IoMT), Cybersecurity, Intrusion Detection, Machine Learning, Data Leakage, Conformal Prediction, Healthcare Security.