A New Method for Heart Disease Detection: Long Short-Term Feature Extraction from Heart Sound Data

Abstract

Heart sounds have been extensively studied for heart disease diagnosis for several decades. Traditional machine learning algorithms applied in the literature have typically partitioned heart sounds into small windows and employed feature extraction methods to classify samples. However, as there is no optimal window length that can effectively represent the entire signal, windows may not provide a sufficient representation of the underlying data. To address this issue, this study proposes a novel approach that integrates window-based features with features extracted from the entire signal, thereby improving the overall accuracy of traditional machine learning algorithms. Specifically, feature extraction is carried out using two different time scales. Short-term features are computed from five-second fragments of heart sound instances, whereas long-term features are extracted from the entire signal. The long-term features are combined with the short-term features to create a feature pool known as long short-term features, which is then employed for classification. To evaluate the performance of the proposed method, various traditional machine learning algorithms with various models are applied to the PhysioNet/CinC Challenge 2016 dataset, which is a collection of diverse heart sound data. The experimental results demonstrate that the proposed feature extraction approach increases the accuracy of heart disease diagnosis by nearly 10%.

Keywords: auscultation; feature selection; heart abnormalities; heart sound classification; long short-term features; machine learning.

Figure 1

Fundamental heart sounds, cardiac cycle, and systolic and diastolic periods.

Figure 2

Heart sound graph of a healthy individual (upper). Heart sound graph of an unhealthy individual (lower).

Figure 3

The classical method of heart sound classification.

Figure 4

Flowchart of the proposed method.

Figure 5

Peaks with values above the threshold (upper). Peaks after extra peak rejection process (lower).

Figure 6

Information levels of the features. Dotted line is used to display the threshold level. Blue squares are the ones below threshold, and red lines are the ones above threshold.