Knowledge Graphs for Clinical Data: Strategies and Challenges with an Application to MIMIC III Dataset

The rapid advancement of digital technologies in healthcare has revealed the urgent need for strategies that improve data integration and interoperability among diverse information sources. This proposal describes a structured approach for representing electronic health records through a knowledge-based model, which facilitates the integration and linking of data across different systems. As an application, it focuses on transforming the MIMIC-III database into RDF (Resource Description Framework), addressing the complexities of converting the dataset into a structured, interoperable format, and discussing different integrity and consistency issues.