Enhancing Security in the Internet of Medical Things: A Machine Learning Approach for Intrusion Detection Systems

Abstract

The Internet of Medical Things (IoMT) revolutionizes healthcare by enabling real-time monitoring and remote care but introduces significant security vulnerabilities. This thesis proposes an intrusion detection system (IDS) using the XGBoost algorithm to address these vulnerabilities. Utilizing the WUSTL-EHMS-2020 dataset, which integrates network flow metrics and patient biometric data, our model demonstrates superior performance with 99.11% accuracy, 98.05% recall, and a prediction time of 0.02 seconds. Comparative analysis with other machine learning models underscores the effectiveness of our approach. The results highlight the potential of ensemble learning methods in enhancing IoMT security, ensuring the protection of sensitive medical data and patient safety in interconnected healthcare environments