linking-umls-concepts

Map OpenMed clinical entities to UMLS Concept Unique Identifiers via the UTS API.

5.0k|615|Updated Oct 4, 2025
One-click install
npx skills add https://github.com/maziyarpanahi/openmed --skill linking-umls-concepts
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: linking-umls-concepts
Source: https://github.com/maziyarpanahi/openmed/tree/main/skills/linking-umls-concepts
Command: npx skills add https://github.com/maziyarpanahi/openmed --skill linking-umls-concepts

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill resolves clinical entities extracted by OpenMed to UMLS Concept Unique Identifiers, enabling consistent concept normalization without bundling or caching license-restricted Metathesaurus data.

Core Features & Use Cases

  • CUI Linking: Search the UMLS Metathesaurus through the user's UTS API key and map disease, drug, chemical, and anatomy mentions to candidate CUIs.
  • Semantic-Type Filtering: Use UMLS semantic types to distinguish clinically relevant concepts and reduce ambiguous or incorrect matches.
  • Cross-Vocabulary Mapping: Cross-walk CUIs to SNOMED CT, ICD-10-CM, RxNorm, MeSH, and other source vocabularies for downstream coding and interoperability.
  • Traceable Clinical NLP: Preserve OpenMed entity labels, confidence scores, and source offsets while storing only permitted identifiers and codes.
  • Use Case: Normalize mentions such as myocardial infarction, heart attack, and MI to one CUI, then retrieve corresponding codes for a clinical data integration workflow.

Quick Start

Use the linking-umls-concepts skill with your UTS_API_KEY to resolve OpenMed-extracted entities to CUIs, filter candidates by semantic type, and cross-walk the selected concepts to the target vocabulary.

Frequently Asked Questions about linking-umls-concepts

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I normalize clinical terms across different vocabularies like SNOMED CT and ICD-10-CM?

Concept normalization maps varied clinical mentions, such as heart attack and MI, to a single UMLS Concept Unique Identifier. This Skill resolves extracted entities through the UMLS Metathesaurus to standardize terminology across different biomedical vocabularies.

How do I map extracted clinical entities to UMLS CUIs using a UTS API key?

You map clinical entities to UMLS CUIs by providing your UTS API key for out-of-process UMLS REST API access. The Skill searches the Metathesaurus to resolve disease, drug, and anatomy mentions to candidate CUIs while preserving original entity labels and confidence scores.

Do I need a UMLS license to crosswalk CUIs to RxNorm and MeSH?

Yes, you need a valid UMLS license and UTS API key to crosswalk CUIs to RxNorm and MeSH. The Skill requires out-of-process UMLS REST API access and strictly complies with UMLS licensing restrictions by storing only permitted identifiers without caching Metathesaurus data.

How does semantic type filtering reduce ambiguous UMLS concept matches?

Semantic type filtering reduces ambiguous UMLS concept matches by applying UMLS semantic types to distinguish clinically relevant concepts. This validation step ensures only appropriate disease, drug, chemical, or anatomy matches are selected during the concept normalization workflow.

What are the limitations of using UMLS REST API for concept normalization?

Limitations include the strict dependency on a user-provided UTS API key and compliance with UMLS licensing restrictions. The Skill cannot bundle or cache license-restricted Metathesaurus data, requiring out-of-process REST API access for every concept normalization and crosswalk operation.