Machine Learning Based Anomaly Detection and Homomorphic Encryption for Securing Electronic Health Records in IoT-Enabled Hospitals

Internet of Things (IoT) devices are used by hospitals to enhance patient care. Such gadgets gather health information and keep it in Electronic Health Records (EHRs). EHR is highly sensitive information that should be secured. Nonetheless, IoT systems raise security and privacy threats. The threats to hospitals are numerous as they include data breaches, insider abuse, and ransomware attacks. These new risks cannot be dealt with using the traditional security methods. Machine learning (ML) can contribute to that by identifying abnormal or aberrant behavior within hospitals. Homomorphic encryption (HE) is one of the techniques that can be used to secure the data and perform the calculation on the encrypted data without disclosing it. The combination of ML and HE can enhance security and privacy of healthcare systems. The current review examines the available literature on ML-based anomaly detection and homomorphic encryption in securing EHRs of hospitals through IoT solutions. It outlines the existing practices, citing their shortcomings, and determining where the research remains unfinished and where it should go in the future.