Abstract
An earlier version (2.0) of the case-based reasoning (CBR) tool, called IDEAS for ICU's, allowed users to compare the ten closest matching cases to the newest patient admission, using a large database of intensive care patient records, and physician-selected matching-weights [1,2]. The new version incorporates matching-weights, which have been determined quantitatively. A faster CBR matching engine has also been incorporated into the new CBR. In a second approach, a back-propagation, feed-forward artificial neural network estimated two classes of the outcome "duration of artificial ventilation" for a subset of the database used for the CBR work. Weight-elimination was successfully applied to reduce the number of input variables and speed-up the estimation of outcomes. New experiments examined the impact of using a different number of input variables on the performance of the ANN, measured by correct classification rates (CCR) and the Average Squared Error (ASE).
Full text
PDF
Selected References
These references are in PubMed. This may not be the complete list of references from this article.
- Féry-Lemonnier E., Landais P., Loirat P., Kleinknecht D., Brivet F. Evaluation of severity scoring systems in ICUs--translation, conversion and definition ambiguities as a source of inter-observer variability in Apache II, SAPS and OSF. Intensive Care Med. 1995 Apr;21(4):356–360. doi: 10.1007/BF01705416. [DOI] [PubMed] [Google Scholar]
- Tu J. V., Guerriere M. R. Use of a neural network as a predictive instrument for length of stay in the intensive care unit following cardiac surgery. Comput Biomed Res. 1993 Jun;26(3):220–229. doi: 10.1006/cbmr.1993.1015. [DOI] [PubMed] [Google Scholar]