Related Experiment Videos
Three types of IS-A statement in diagnostic classifications: three types of knowledge needed for development and
F J Flier1, P F de Vries Robbé, P E Zanstra
1Department of Medical Informatics Epidemiology and Statistics, University of Nijmegen, The Netherlands. P.Zanstra@mie.kun.nl
Methods of Information in Medicine
|December 29, 1998
Summary
Updating diagnostic classifications requires understanding different knowledge types and IS-A statements. Collaboration between medical terminologists and scientists ensures accurate revisions and comparability across classification versions.
Area of Science:
- Medical Informatics
- Medical Terminology
- Knowledge Representation
Background:
- Diagnostic classifications require robust update mechanisms to reflect evolving medical knowledge.
- Maintaining comparability across different versions of classifications is crucial for consistent data use.
- Existing update mechanisms may oversimplify the complexities of medical knowledge integration.
Purpose of the Study:
- To provide a framework for updating diagnostic classifications.
- To describe different types of IS-A statements and their associated knowledge.
- To facilitate collaboration between medical terminologists and scientists for classification maintenance.
Main Methods:
- Analysis of IS-A (is-a) statement types: 'by definition', 'probably', and 'in theory necessarily'.
- Categorization of knowledge types: linguistic conventions, probabilities, and empirical theories.
- Examination of the role of medical scientists in updating classifications.
Main Results:
- Three distinct types of IS-A statements correspond to different knowledge bases.
- 'A is by definition a B' relies on linguistic conventions.
- 'A is probably a B' uses probabilistic knowledge.
- 'A is in theory necessarily a B' involves empirical theories and their status.
- Updating necessitates considering changes in probabilities and empirical theories.
Conclusions:
- Diagnostic classification development and maintenance require interdisciplinary collaboration.
- Medical scientists are essential for updating classifications due to evolving medical knowledge.
- The linguistic concept of hyponymy is insufficient for complex diagnostic classification updates.