D-CIS Publication Database


Type of publication:Inproceedings
Entered by:JOSM
TitleInference in the Promedas medical expert system
Bibtex cite ID
Booktitle Proceedings of the 11th Conference on Artificial Intelligence in Medicine (AIME 2007)
Year published 2008
Month July
Publisher Springer
Location 7-11 July 2007, Amsterdam, The Netherlands
Note ISBN 978-3-540-73598-4
Keywords inference,promedas,medical,expert system
In the current paper, the Promedas model for internal medicine, developed by our team, is introduced. The model is based on up-to-date medical knowledge and consists of approximately 2000 diagnoses, 1000 ndings and 8600 connections between diagnoses and ndings, covering large parts of internal medicine. Promedas is currently being evaluated informally by specialists in internal medicine at the Utrecht university hospital and is receiving positive responses.We show that Belief Propagation (BP) can be successfully applied to approximate inference problems in the Promedas network. BP converges on all patient test cases, which were generated with the help of the model itself. In some cases, however, we nd errors that are too large for this application. We apply a recently developed method that improves the BP results by means of a loop expansion scheme. This method, termed Loop Corrected (LC) BP, is able to improve the marginal probabilities signi cantly, leaving a remaining error which is acceptable for the purpose of medical diagnosis.
Wemmenhove, Bastian
Mooij, Joris
Wiegerinck, Wim
Leisink, Martijn
Kappen, Hilbert J.
Total mark: 5