Related Experiment Videos
The findings--diagnosis continuum: implications for image descriptions and clinical databases
R A Greenes1, R C McClure, E Pattison-Gordon
1Decision Systems Group, Harvard Medical School, Brigham and Women's Hospital, Boston, Massachusetts.
Proceedings. Symposium on Computer Applications in Medical Care
|January 1, 1992
Summary
This study introduces a new semantic net model for medical ontologies. It enhances the Unified Medical Language System (UMLS) by improving clinical finding representation and adaptable data retrieval for diverse user needs.
Area of Science:
- Medical Informatics
- Knowledge Representation
- Ontology Engineering
Background:
- The Unified Medical Language System (UMLS) Metathesaurus faces challenges in representing clinical findings due to varied descriptions and granularity.
- Existing systems struggle with indexing and retrieving information across diverse databases with unique vocabularies.
Purpose of the Study:
- To develop an adaptable medical ontology using semantic net representation.
- To address deficiencies in the UMLS Metathesaurus, particularly concerning clinical findings.
- To facilitate cross-resource retrieval by maintaining a consistent concept representation.
Main Methods:
- Exploration of semantic net representation within the UMLS project.
- Development of a recursive model for representing observations and interpretations.
- Focus on a continuum of aggregation for findings, from perceptual to interpretive levels.
Main Results:
- A recursive semantic net model was developed for observations and interpretations.
- The model demonstrates adaptability to varying perspectives and aggregation levels of clinical findings.
- Potential for improved data indexing and retrieval across heterogeneous databases.
Conclusions:
- The proposed semantic net model offers a flexible approach to medical ontology construction.
- This method enhances the representation of clinical findings, addressing limitations in current systems.
- The model facilitates more comprehensive and adaptable information retrieval for diverse users and purposes.