Related Experiment Videos
SCOUT: information retrieval from full-text medical literature.
1Section on Medical Informatics, Stanford University School of Medicine, California 94305-5479.
Proceedings. Symposium on Computer Applications in Medical Care
|January 1, 1992
Summary
SCOUT is an expert system helping physicians find oncology information in biomedical literature. It uses expert knowledge to retrieve only the most relevant documents for patient-specific questions.
Area of Science:
- Medical Informatics
- Oncology
- Information Retrieval
Background:
- Practicing physicians require efficient access to biomedical literature for patient care.
- Existing information retrieval systems may return excessive irrelevant documents.
- Oncology literature search demands specialized knowledge for accurate results.
Purpose of the Study:
- To introduce SCOUT, an expert system for oncology literature searching.
- To improve the retrieval of patient-specific information for clinicians.
- To minimize the number of documents returned while maximizing relevance.
Main Methods:
- SCOUT models the expertise of oncologists and information specialists.
- It employs a probabilistic inference system.
- Expert heuristics are used for source selection and search statement development.
Main Results:
- SCOUT effectively identifies optimal information sources for clinical queries.
- It generates source-specific search statements.
- The system retrieves a minimal number of pertinent documents.
Conclusions:
- SCOUT enhances the efficiency of searching biomedical literature in oncology.
- The system provides a valuable tool for clinical decision support.
- It successfully addresses the need for precise information retrieval in patient care.