Detection of Heart Sound Murmurs and Clinical Outcome with Bidirectional Long Short-Term Memory Networks

Abstract

Heart sound recordings are a key non-invasive tool to detect both congenital and acquired heart conditions. As part of the George B. Moody PhysioNet Challenge 2022, we present an approach based on Bidirectional Long Short-Term Memory (BiLSTM) neural networks for the detection of murmurs and prediction of clinical outcome from Phonocardiograms (PCGs). We used the homomorphic, Hilbert, power spectral density, and wavelet envelopes as signal features, from which we extracted fixed-length segments of 4 seconds to train the network. Using the official challenge scoring metrics, our team SmartBeatIT achieved a murmur weighted accuracy score of 0.757 on the hidden test set (ranked 6th out of 40 teams), and an outcome cost score of 13815 (ranked 25th out of 39 teams). With 5-fold cross-validation on the training set, in the murmur detection task we obtained sensitivities of 0.827 and 0.312 for the Present and Unknown classes and a specificity of 0.801; and a sensitivity of 0.676 and a specificity of 0.544 in the outcome prediction task.