Using SNOMED semantic concept groupings to enhance semantic-type assignment consistency in the UMLS

Authors:
Duo Wei;Michael Halper;Gai Elhanan
Affiliations:
The Richard Stockton College of NJ, Galloway, NJ, USA;New Jersey Institute of Technology, Newark, NJ, USA;Halfpenny Technologies, Inc., Blue Bell, PA, USA
Venue:
Proceedings of the 2nd ACM SIGHIT International Health Informatics Symposium
Year:
2012

Citing 6
Cited 0

Structural group auditing of a UMLS semantic type's extent

Journal of Biomedical Informatics
A review of auditing methods applied to the content of controlled biomedical terminologies

Journal of Biomedical Informatics
Analyzing polysemous concepts from a clinical perspective: Application to auditing concept categorization in the UMLS

Journal of Biomedical Informatics
The Neighborhood Auditing Tool: A hybrid interface for auditing the UMLS

Journal of Biomedical Informatics
Alignment of the UMLS semantic network with BioTop

Bioinformatics
Auditing concept categorizations in the UMLS

Artificial Intelligence in Medicine

Quantified Score

Hi-index	0.00

Visualization

Abstract

SNOMED concepts make up a significant percentage of the UMLS Metathesaurus. Hence, the correctness of their semantic type (ST) assignments contributes significantly to the overall correctness of ST assignments throughout the UMLS. Two kinds of semantic concept groupings within SNOMED are employed to automatically check for potentially inconsistent ST assignments, based on their collective sets of assignments. The first kind of concept grouping, called a semantic uniformity group (SUG), is based on concepts' properties and hierarchical configurations. The second kind constitutes intersections of groupings of the first kind ("overlapping SUG" (OSUG)). The methodology is applied to SNOMED's Specimen hierarchy, where 448 concepts residing in 40 SUGs are analyzed. A hypothesis concerning the fact that OSUGs can be stronger indicators of inconsistencies than SUGs is considered. The results show that our methodology can be effective in finding ST assignment problems and thus can augment the suite of techniques available for managing the enormous amount of such assignments.