Extension of the HEPAR II model to multiple-disorder diagnosis

Authors:

Agnieszka Onisko
Bialystok University of Technology
Institute of Computer Science
Bialystok, 15-351, Poland
e-mail: aonisko@ii.pb.bialystok.pl
FAX: (085) 422-393

Marek J. Druzdzel
Decision Systems Laboratory
School of Information Sciences
and Intelligent Systems Program
University of Pittsburgh
e-mail: marek@sis.pitt.edu

Hanna Wasyluk
The Medical Center of Postgraduate Education
and Institute of Biocybernetics
and Biomedical Engineering,
Polish Academy of Sciences
Warsaw, Marymoncka 99, Poland
e-mail: hwasyluk@cmkp.edu.pl

Abstract:

The HEPAR II system is based on a Bayesian network model of a subset of the domain of hepatology in which the structure of the network is elicited from an expert diagnostician and the parameters are learned from a database of medical cases. The model follows the assumption made in the database that each patient case is diagnosed with a single disorder, i.e., disorders are mutually exclusive.

In this paper, we describe an extension of the HEPAR II system to multiple-disorder diagnosis. We show that our network transforms readily to a network that can perform multiple-disorder diagnosis with some benefits to the quality of numerical parameters learned from the database. We demonstrate empirically that the diagnostic performance in terms of single-disorder diagnosis improves under this transformation. The new model is more realistic and we expect that it will be of higher value in clinical practice.