Entity Type Search

After the NLP API extracts and maps a recognized medical term to concepts in supported ontologies, it also performs post-coordination to map the term (or more correctly the entity which represents an instance of the term), to semantic types defined in these ontology. For example, the term diabetes has the following semantic types in the respective ontologies:

• SNOMED: disorder
• Medcin: diagnosis
• Radlex: clinical finding
• UMLS: Disease or Syndrome.

In the NLP API, semantic types are referred to as entity types, and stored in the foundentitytype table. The post-coordination of concepts to entity types provides maximum flexibility for data retrieval, allowing for example the user to write queries which returns recognized terms as represented in one ontology (e.g. SNOMED CT) , but are classified in another (e.g. using UMLS semantic types).

This section provides examples of query patterns that use entity types as a search parameter. Entity type searches are especially useful when you need to do a broad-based search for data exploration or summarization, for example.