Graph and Structured Data Algorithms in Electronic Health Records: A Scoping Review

The provision of some of the most fundamental services related to health and wellbeing makes healthcare an essential part of society. Patient information management and accessibility are only one of many facets of healthcare that have changed as a result of technological improvements over time. Electronic health records (EHRs) being differentiated as an emerging integral component of contemporary healthcare systems, provide digital methods for patient data storage and organization. As the volume of EHRs continues to increase exponentially, it is becoming increasingly important to optimize the reasoning process in this enormous amount of data. This scoping review aims to provide a comprehensive and systematic comparison between graph and non-graph structured data in regards to their efficiency in reasoning regarding EHRs. Research Articles from prominent research paper repositories such as Science Direct, IEEExplore, SpringerLink, and ACM are analyzed in detail. Their corresponding techniques and algorithms are gathered, analyzed and concluded that neural networks are mostly used in both graph and structured data, providing a comparison of the best reasoning techniques between aforementioned data structures.