Topological-Pattern-Based Recommendation of UMLS Concepts for National Cancer Institute Thesaurus

Area of Science:

• Medical Informatics
• Bioinformatics
• Ontology Development

Background:

• The National Cancer Institute Thesaurus (NCIt) is a vital reference terminology for cancer research and administration.
• Medical knowledge is constantly evolving, necessitating regular updates and additions to terminologies like NCIt.
• Manual terminology development is labor-intensive and prone to errors, highlighting the need for efficient, semi-automated methods.

Purpose of the Study:

• To evaluate a semi-automated method for identifying potentially missing concepts to enrich the National Cancer Institute Thesaurus (NCIt).
• To adapt a previously successful structural method, leveraging the Unified Medical Language System (UMLS) term mappings, for NCIt enrichment.

Main Methods:

• A structural method was employed, utilizing native term mappings from the Unified Medical Language System (UMLS).
• Concepts from eight distinct UMLS source terminologies were analyzed for their potential to augment NCIt's existing conceptual framework.

Main Results:

• The analogous method successfully identified potential concepts from UMLS source terminologies for NCIt enrichment.
• The study demonstrated the feasibility of using UMLS mappings to find candidate concepts for NCIt.

Conclusions:

• The developed semi-automated method offers a cost-effective approach for terminology curators to identify and incorporate missing concepts into the NCIt.
• This strategy aids in keeping the NCIt current and comprehensive, supporting advancements in cancer research and clinical practice.