Machine Learning Approaches to Mitigate Insider Threats in Electronic Health Records Systems

The increasing of convert the healthcare system to be as digital system, becoming necessary to secure electronic health records against inside threats. Although the existing numerous research provides solutions and utilizes machine learning techniques to enhance security of the records, it remains fragmented and lacks comprehensive synthesis of approaches. In this paper aims to present a comprehensive and rigorous study for an appropriate method based on ML to detect and mitigate insider threats in EHR systems. PRISMA methodology is used to follow. It starts to analyze 537 primary studies from different major databases where it looks to identify the relevant research in the same scope. Then it reduced to 25 studies after applying the standards of inclusion and exclusion. This study focuses on comparing the algorithms and techniques that have been used, the challenges faced with the implementation process, and the gap of existing studies. The study findings reveal trends, limitations, challenges, and future directions to develop and support the intelligence system and protect privacy and secure healthcare records.