There are three different models that make up the knowledge base of the Heart Failure Program. Two of these are integrated and form a static physiologic knowledge base and the third is built dynamically from cases. The first is the diagnostic causal model. This model consists of pathophysiologic states connected by probabilities representing the likelihood of causation. The second is the quantitative physiologic parameter model used for therapy prediction. It consists of parameters related by equations representing the constraints between the parameters. The final model is the case base of completed diagnoses built by the case based reasoner and organized for matching against new cases. It is also used for diagnosis and offers an alternate way of producing a diagnosis.
The following sections give a description of each of these models and highlight the relationships that bind them into a unified whole.