Conference presentation
Representing medical decision making strategies in a CBR system
Drexel University. College of Information Science and Technology. Faculty Publications and Research.
04 Mar 2008
Abstract
This paper describes and compares the development of two organizational structures to represent medical decision making strategies. We generate the solution to a new problem by applying a previous solution from a medical record in a CBR system that performs decision-making about hypertension drug therapy. The case libraries are structured in accordance with the approaches of flat memory and discrimination network. Cases are originated by a retrospective knowledge acquisition about 47 patients who underwent ambulatory care of a university hospital. The similarity-based retrieval employed in the flat structure resembles what physicians do when handling their routine cases of arterial hypertension. Physicians identify a similar case in memory by recognizing the content embedded in the new situation, like a script. The hypothetico-deductive method for searching the case solution follows a similar strategy to the one represented in the prioritized discrimination network. The inclusion of cases in the case library of the discrimination network required more complex procedures than in the case library of the flat memory. These two decision support systems could contribute significantly to patient care. The system we are researching on has educational purposes as well.
Metrics
7 File views/ downloads
1 Record Views
Details
- Title
- Representing medical decision making strategies in a CBR system
- Creators
- Roberto Henrique Heinisch (Author) - Drexel University (1970-)Rosina O. Weber (Author) - Drexel University (1970-)Alejandro Martins (Author) - Drexel University (1970-)Ricardo Miranda Barcia (Author) - Drexel University (1970-)
- Publication Details
- Drexel University. College of Information Science and Technology. Faculty Publications and Research.
- Resource Type
- Conference presentation
- Language
- English
- Academic Unit
- DU; College of Information Science and Technology (1995-2013)
- Identifiers
- 991014632714704721