A prototype ETL pipeline that uses HL7 FHIR RDF resources when deploying pure functions to enrich knowledge graph...

Area of Science:

• Biomedical Informatics
• Data Science
• Clinical Research Informatics

Background:

• Knowledge graphs (KGs) are valuable for clinical care and research.
• Enriching KGs with patient data requires systematic and transparent methods.
• Leveraging existing ontologies and data standards promotes semantic interoperability.

Purpose of the Study:

• To develop and test a novel data processing pipeline for KG enrichment.
• To enable the computation and integration of new patient features into existing KGs.
• To demonstrate the use of standards-based methods for transparent data processing.

Main Methods:

• Developed a data processing pipeline to extract, compute, and return new patient features.
• Utilized electronic health record and patient survey data within a large KG.
• Employed Health Level 7's FHIR RDF standards for data representation and validation.

Main Results:

• Successfully implemented and tested a pipeline for enriching patient KGs.
• Demonstrated programmatic validation of RDF data resource types specified by FHIR.
• Showcased the pipeline's capability to represent newly derived patient-level features.

Conclusions:

• Knowledge graph technology can be enhanced with standards-based semantic data processing pipelines.
• Pure functions can be deployed and traced for deriving new patient features.
• Semantic data processing pipelines facilitate reporting on new patient computations with linked metadata.