Map source clinical concepts to OMOP standards via DuckDB search.
npx skills add https://github.com/BorisDelange/Linkr --skill concept-mapping-ai
Please help me install this Agent Skill. Skill: concept-mapping-ai Source: https://github.com/BorisDelange/Linkr/tree/main/.claude/skills/concept-mapping-ai Command: npx skills add https://github.com/BorisDelange/Linkr --skill concept-mapping-ai
It reduces the manual effort and inconsistency of mapping source clinical concepts (measurements, conditions, procedures, observations) to the correct OMOP standard concepts by generating candidate matches and providing structured reasoning.
similarity-scores.parquet or as reviewer-confirmed authored mappings in mappings.json.Ask the concept-mapping orchestrator to run the concept-mapping-ai skill for your selected source concept batch with an existing DuckDB session context and a chosen destination mode (suggestions or mappings).
High-intent search queries and answers about installing and using this skill.
Mapping source clinical concepts to OMOP standard concepts is done by generating and ranking candidate matches using DuckDB search, precomputed similarity scores, and clinical reasoning across batches. The Skill assigns SSSOM equivalence levels with evidence comments to support reviewer decisions.
Yes, DuckDB is used to search loaded OMOP tables for concept mapping candidates. A prepared DuckDB session context with defined OMOP tables is required to generate ranked matches for source measurements, conditions, procedures, and observations.
SSSOM equivalence levels are assigned by evaluating generated candidate matches and applying clinical reasoning to grade them as exactMatch, closeMatch, broadMatch, narrowMatch, or relatedMatch. Evidence-oriented comments are provided for each graded mapping to support reviewer decisions.
Batch processing with per-batch review modes is supported for mapping clinical concepts to OMOP. The Skill operates on prepared DuckDB session context and handles edge cases like missing matches or non-standard concepts across selected source concept batches.
OMOP concept mapping suggestions can be output as rows formatted for a similarity parquet file for UI review, or as approved ConceptMapping objects written to a mappings.json file for reviewer-confirmed authored mappings.
Automating clinical reasoning for OMOP concept mapping is best handled by an agentic workflow that generates ranked candidates using DuckDB search and precomputed similarity scores. It reduces manual effort and inconsistency by applying structured reasoning across batches.
BorisDelangeMap source drug concepts to OMOP RxNorm Clinical Drug concepts.
BorisDelangeOrchestrate OMOP concept mapping with syntactic and semantic candidate scoring.
kriegcloudGenerate and validate SSSOM mapping sets between ontologies using lexical matching and LLM-assisted verification.
indicate-euGenerate detailed Markdown descriptions of clinical concept sets using biomedical vocabularies.