A Prototype ETL Pipeline Using HL7 FHIR RDF Standards for Knowledge Graph Data Enrichment

Details

The research paper, published in the Journal of Biomedical Semantics, is titled 'A prototype ETL pipeline that uses HL7 FHIR RDF resources when deploying pure functions to enrich knowledge graph patient data.' The core objective of this prototype ETL pipeline is to enrich knowledge graph patient data by leveraging HL7 FHIR resources and Resource Description Framework (RDF). Specifically, the study focuses on enhancing the process of converting Electronic Health Record (EHR) data into a knowledge graph using pure functions. This approach utilizes mechanisms involving FHIR RDF Library, Observation, and Provenance resources to handle complex data structures. Technically, the research integrates multiple tools and frameworks, including Apache NiFi, Python's PHQ-9 scoring API, HL7 FHIR RDF Library (Observation, Provenance), ShEx.js for RDF validation, and CAMH’s Blue Brain Nexus. This combination enables a complete workflow from data collection through ETL processing to knowledge graph storage. This work demonstrates that combining the FHIR standard with advanced semantics like RDF moves beyond simple data migration (ETL). It aims to build knowledge graphs with semantic richness. In the context of Japanese healthcare IT, this represents an advanced model for data integration and utilization, suggesting that FHIR can serve as a foundational base not just for data format but also for sophisticated semantic processing when combined with RDF.