Related Experiment Videos
Automated knowledge extraction for decision model construction: a data mining approach.
Ai -Ling Zhu1, Jian Li, Tze -Yun Leong
1School of Computing, National University of Singapore, 117543.
AMIA ... Annual Symposium Proceedings. AMIA Symposium
|January 20, 2004
Summary
This study presents an automated method to extract medical term relationships from MEDLINE citations, aiding clinical decision model construction. The approach successfully identified causal relations and decision alternatives in colorectal cancer management.
Area of Science:
- Medical Informatics
- Natural Language Processing
- Biomedical Data Mining
Background:
- Medical Subject Headings (MeSH) and Subheadings in MEDLINE citations offer potential for inferring relationships between medical concepts.
- Automated extraction of these semantic relations is crucial for developing clinical decision support systems.
Purpose of the Study:
- To propose and evaluate an automated approach for extracting semantic relations among medical terms from MEDLINE citations.
- To facilitate the construction of clinical decision models by deriving decision elements and their relationships.
Main Methods:
- Utilizing the Apriori association rule mining algorithm to identify co-occurrences of medical concepts.
- Filtering concept co-occurrences through predefined semantic templates to establish meaningful relations.
Main Results:
- Demonstrated the extraction of useful causal relations from MEDLINE data.
- Successfully identified decision alternatives relevant to clinical management.
Conclusions:
- The proposed method effectively extracts semantic relations from biomedical literature.
- This approach supports the development of data-driven clinical decision models, as shown in a colorectal cancer case study.