Enhancing Cybersecurity Threat Detection and Mitigation in Telemedicine Using Artificial Intelligence

The expansion of telemedicine over the last few years has greatly increased access to care in otherwise underserved geographic areas and in areas with very limited resources. However, this shift toward digitization has also been accompanied by tremendous risk to cybersecurity that can compromise confidential patient information’s confidentiality, integrity, and availability. The goal of this study is to evaluate the application of Artificial Intelligence (AI) techniques, that is, Machine Learning (ML) and Deep Learning (DL), in enhancing the cybersecurity threat detection and response of telemedicine systems. We tested the two new AI models, Random Forest (RF), and Long Short-Term Memory (LSTM), with the CICIDS2017 dataset to determine their capability in Internet of Medical Things (IoMT) devices-specific network intrusion detection. The RF model did not disappoint, way above expectation by recording 99.99% detection accuracy, and barely any latency. The LSTM model, although having a slightly lower accuracy of 96.9%, was hugely effective in handling the temporal characteristics of data as well as real-time anomaly detection. The performance demonstrated that both models significantly enhance intrusion detection ability and can be adjusted to be compatible with different configurations of telemedicine networks. Moreover, this research presents the ability of AI to solve the issue of data imbalance, ensure privacy, and provide dynamic solutions for changing healthcare settings.