Related Experiment Videos
Application of a network structure in a knowledge-based system for medical diagnostics.
1Clinical Neurophysiology, University Hospital, Göttingen, F.R.G.
Computer Methods and Programs in Biomedicine
|November 1, 1990
Summary
This study introduces a novel semantic net approach for diagnostic decision-making systems. The developed expert system, PROJECTOR, demonstrates efficient data handling for differential diagnosis on personal computers.
Area of Science:
- Computer Science
- Medical Informatics
- Artificial Intelligence
Background:
- Diagnostic decision-making systems traditionally rely on production rules.
- Limited conceptual frameworks exist for processing information using semantic nets.
- This research addresses the need for advanced semantic net applications in diagnostics.
Purpose of the Study:
- To describe the application of a semantic net for diagnostic decision support.
- To introduce a method for defining information sets using semantic net views.
- To present the PROJECTOR expert system for differential diagnosis.
Main Methods:
- Implementation of a semantic net using a network-like pointer structure in Pascal.
- Creation of various 'views' to define context-specific information sets.
- Utilization of 'projections' (mappings and set operations) for information inference.
Main Results:
- The developed semantic net structure is fast and compact.
- The system efficiently handles large datasets.
- The PROJECTOR expert system demonstrates feasibility for personal computer use.
Conclusions:
- Semantic nets offer a viable and efficient alternative to production rules for diagnostic systems.
- The developed projection method effectively infers information within the semantic net.
- The PROJECTOR system showcases the practical application of semantic nets in differential diagnosis.