Development of artificial neural networks based algorithms for the classification of bileaflet mechanical heart valve sounds

Purpose: As all mechanical prostheses, bileaflet heart valves are prone to thrombus formation; reduced hemodynamic performances and embolic events can hence occur. Prosthetic valve thrombosis affects the power spectra calculated from the phonocardiographic signals corresponding to prostheses closing events. Artificial neural network-based classifiers are proposed for automatically and non-invasively assessing valve functionality and detecting thrombotic formations. Further studies will allow enlarging data set, extending the investigated frequency range, and applying the presented approach to other bileaflet mechanical valves. Methods: Data were acquired for the normofunctioning St. Jude Regent valve mounted in the aortic position of a Sheffield Pulse Duplicator. Different pulsatile flow conditions were reproduced, changing heart rate and stroke volume. The case of a thrombus completely blocking one leaflet was also investigated. Power spectra were calculated from the phonocardiographic signals and used to train artificial neural networks of different topologies; neural networks were then tested with the spectra acquired in vivo from 33 patients, all recipients of the St. Jude Regent valve in the aortic position. Results: The proposed classifier showed 100% correct classification in vitro and 97% when applied to in vivo data: 31 spectra were assigned to the right class, 1 was classified as false positive, and 1 as “not classifiable”. Conclusion: Early malfunction detection is necessary to prevent thrombotic events on bileaflet mechanical heart valves. Following further clinical validation on extended patients’ database, artificial neural network-based classifiers could be embedded in a portable device able to detect valvular thrombosis at early stages of formation: it would help clinicians in formulating valvular dysfunction diagnosis before the appearance of critical symptoms.