Probabilistic Neural Networks with Macrolayers in Systems for Support of Decision-Making in Differential Diagnosis of Cardiovascular Diseases

Probabilistic neural networks with macrolayers are used in systems for support of decision-making in differential diagnosis of cardiovascular diseases. The neural network contains two macrolayers. Macrolayer 1 is based on the Bayesian probabilistic formalism. Macrolayer 2 is based on uncertain logic of decision-making and provides classification into two classes. The neural network provides differential diagnosis of cardiovascular diseases. The subjective probabilistic factors are estimated in points or percents. The expert estimates the probability of the zero hypothesis on the basis of a posteriori probability of the alternative hypothesis.