Automatic Feature Extraction for Heartbeat Anomaly Detection

TL;DR

This paper introduces an unsupervised autoencoder-based method for extracting features from raw heartbeat sounds to detect anomalies, utilizing advanced probabilistic modeling and testing on a standard healthcare dataset.

Contribution

It proposes a novel autoencoder architecture with a variational inference objective and Gaussian chain model to capture temporal correlations in heartbeat sounds.

Findings

Effective feature extraction for heartbeat anomaly detection.

Improved detection performance over existing methods.

Demonstrated on PASCAL Heart Sounds Challenge dataset.

Abstract

We focus on automatic feature extraction for raw audio heartbeat sounds, aimed at anomaly detection applications in healthcare. We learn features with the help of an autoencoder composed by a 1D non-causal convolutional encoder and a WaveNet decoder trained with a modified objective based on variational inference, employing the Maximum Mean Discrepancy (MMD). Moreover we model the latent distribution using a Gaussian chain graphical model to capture temporal correlations which characterize the encoded signals. After training the autoencoder on the reconstruction task in a unsupervised manner, we test the significance of the learned latent representations by training an SVM to predict anomalies. We evaluate the methods on a problem proposed by the PASCAL Classifying Heart Sounds Challenge and we compare with results in the literature.