Computer-based diagnostic decision support systems (DSS) will play an increasingly important role in health care. Due to the inherent probabilistic nature of medical diagnosis, a DSS should preferably be based on a probabilistic model. In particular Bayesian networks provide a powerful and conceptually transparent formalism for probabilistic modelling.
A drawback is that Bayesian networks become intractable for exact computation if a large medical domain would be modeled in detail. This has obstructed the development of a useful system for internal medicine. Advances in approximation techniques have opened new possibilities to deal with the computational problem. In this research project, we develop a Bayesian network model for diagnosis of patients in internal medicine. In addition, we develop efficient methods for approximate inference for Bayesian networks.