Anomaly Detection for Internet of Medical Things Using Chameleon Optimization-Based Feature Selection

The Internet of Medical Things (IoMT) offers immense potential for transforming healthcare through connected medical devices and wearables . However, IoMT systems are vulnerable to cyberattacks that can jeopardize patient safety. This article proposes an intrusion detection framework to identify anomalies and attacks in IoMT networks. A chameleon optimization algorithm is utilized for efficient feature selection from network traffic data. The selected features are leveraged to train a classifier to distinguish between normal and abnormal traffic. Experiments demonstrate that the proposed approach with chameleon-based feature selection significantly enhances detection accuracy and reduces false alarms compared to conventional techniques. The intrusion detection model exhibits high attack detection rates while minimizing the number of features required. This research enables robust anomaly detection for IoMT networks, augmenting security and safeguarding patient health through intelligent optimization and machine learning.