OntologyRAG introduces a novel approach that enhances biomedical code mapping by integrating knowledge graphs with retrieval-augmented generation.

• Achieves higher accuracy in mapping biomedical concepts across different ontologies
• Operates significantly faster than traditional methods
• Reduces the manual workload for coding experts by generating higher quality initial mappings
• Leverages the structured knowledge in biomedical ontologies to improve LLM performance

This innovation matters for healthcare organizations by standardizing medical data interpretation, enabling more consistent patient records, and improving interoperability between different medical systems and research databases.

OntologyRAG: Better and Faster Biomedical Code Mapping with Retrieval-Augmented Generation (RAG) Leveraging Ontology Knowledge Graphs and Large Language Models