Classification of Carpentier-Edwards bioprosthesis heart valves using an adaptive single layer perceptron

Presents the application of a single layer perceptron (SLP) for classification of normal (NF) and malfunctioning (MF) Carpentier-Edwards (C-E) bioprosthetic heart valves. Twenty six patients with a C-E valve implanted in the aortic position mere investigated in this study. Three features extracted from the analysis of the second heart sound (S2) by modified forward-backward Prony method (MFBPM) are used as the input vector of an adaptive SLP. Results show that this method gives 100% correct classification between NF and MF cases. However, further investigation is needed to validate the performance of the method in a greater number of patients.