Securing electronic health records against insider-threats: A supervised machine learning approach

The introduction of electronic health records (EHR) has created new opportunities for efficient patient data management. For example, preventative medical practice, rather than reactive, is possible through the integration machine learning to mine digital record datasets. Furthermore, within wider smart cities’ infrastructure, EHR considerable environmental and cost-saving benefits healthcare providers. Yet, there are inherent dangers digitising records. Considering sensitive nature data, equally at risk both external threats insider attacks, but security applications predominantly facing outer boundary network. Therefore, in this work, focus on misuse detection. approach involves use supervised classification (decision tree, random forest support vector machine) based off pre-labelled real-world collated from a UK-based hospital detection misuse. results demonstrate that by employing analyse access, anomaly can be achieved with 0.9896 accuracy test set 0.9908 validation using classifier. emphasis research misuse, anomalous behavioural patterns. Based results, recommendation adopt an SVM misuse/insider threat