Securing healthcare systems and optimizing data analytics through IoMT threat detection

The integration of Internet of Medical Things (IoMT) devices into are systems has raised significant cybersecurity concerns, especially regarding threats to patient safety and the protection of sensitive medical data. This research proposes a novel hybrid machine learning framework aimed at improving the detection and mitigation of cyberattacks in IoMT environments. Our approach combines random forest, AdaBoost, and bagging algorithms to identify various attack vectors across different IoMT networks. We evaluate our framework using a comprehensive IoMT traffic dataset that includes different communication protocols. Using advanced statistical profiling and ensemble classification models, our system achieves high detection performance while significantly reducing false positive rates compared to traditional methods. The hybrid model demonstrates an exceptional precision of 99.92%, ensuring reliable differentiation between benign and malicious network traffic and minimizing disruptions in critical healthcare environments. Experimental validation across various attack scenarios confirms the effectiveness of the framework in addressing the unique security challenges posed by resource-constrained IoMT devices and heterogeneous communication protocols.