Skip to Main Content
Heart failure is a complex syndrome that affects more than 5% of the population over 65, and whose direct costs account for a 2% of the health budget in developed countries. The existing interrelations among the different causes, mechanisms, symptoms and treatments associated to the condition complicate its modeling and, hence, the development of decision support tools which assist health professionals. This article describes the use of a Bayesian network in the modeling of heart contractility dysfunctions reflected in the condition of systolic heart failure and the use of influence diagrams in the decision for treatment actions. The resulting network estimates the probability of a patient for developing an asymptomatic ventricular systolic dysfunction and systolic heart failure from the specification of signs, symptoms, risk factors, cardiovascular disorders or diagnosis tests results. Based on that, the network informs about the convenience of applying a preventive or a corrective treatment.