UMLS
About the resource
The Unified Medical Language System (UMLS), maintained by the NLM, unifies more than 200 biomedical source vocabularies — SNOMED CT, MeSH, ICD-9/10/11, RxNorm, LOINC, the NCI Thesaurus, MedDRA, OMIM, HPO, Orphanet's ORDO and many more — into a single Metathesaurus. Each clinical concept (a CUI, a Concept Unique Identifier) groups synonymous terms across vocabularies and is associated with semantic types from the UMLS Semantic Network.
UMLS is the connective tissue across nearly every disease database in this registry: it makes 'this code in EHR vocabulary X means the same disease as that code in research vocabulary Y' a machine-decidable question. The Metathesaurus is licensed-free for research after a one-time UMLS Terminology Services registration. Releases ship twice yearly.
What you'd use it for
- 01Map an EHR vocabulary code to its research-ontology equivalent
- 02Build a clinical-NLP pipeline grounded in UMLS CUIs
- 03Subset the Metathesaurus to a specific use case via MetamorphoSys
- 04Cross-reference diseases across SNOMED, ICD, MeSH and OMIM