Related Experiment Videos
MEDTAG: tag-like semantics for medical document indexing
1Medical Informatics Division, University Hospital of Geneva, Switzerland.
Proceedings. AMIA Symposium
|November 24, 1999
Summary
Natural Language Processing (NLP) enhances medical documentation by creating a semantic tagset using standard resources like UMLS classes. This improves information retrieval and knowledge extraction from clinical notes.
Area of Science:
- Medical Informatics
- Natural Language Processing
Background:
- Medical documentation is crucial for healthcare communication.
- A gap exists between storing medical information and extracting actionable knowledge.
- Natural Language Processing (NLP) offers a solution for managing large volumes of clinical text.
Purpose of the Study:
- To construct a semantic tagset for indexing medical documents.
- To leverage standard medical resources for tagset development.
- To evaluate the utility of semantic tagging in information retrieval.
Main Methods:
- Developed a semantic tagset using Unified Medical Language System (UMLS) hierarchical classes.
- Utilized standard medical resources to avoid creating a proprietary tagset.
- Assessed the performance of the semantic tagger.
Main Results:
- The semantic tagset aids in disambiguating medical terms.
- Semantic tagging enhances query expansion in retrieval systems.
- The constructed tagger demonstrated effective results in assessments.
Conclusions:
- A UMLS-based semantic tagset is effective for medical document indexing.
- Semantic tagging improves both information retrieval and knowledge extraction.
- NLP-driven semantic analysis is vital for advancing medical informatics.