Abstract

Heart sounds play an important role in the diagnosis of cardiac conditions. Due to the low signal-to-noise ratio (SNR), it is problematic and time-consuming for experts to discriminate different kinds of heart sounds. Thus, objective classification of heart sounds is essential. In this study, we combined a conventional feature engineering method with deep learning algorithms to automatically classify normal and abnormal heart sounds. First, 497 features were extracted from eight domains. Then, we fed these features into the designed convolutional neural network (CNN), in which the fully connected layers that are usually used before the classification layer were replaced with a global average pooling layer to obtain global information about the feature maps and avoid overfitting. Considering the class imbalance, the class weights were set in the loss function during the training process to improve the classification algorithm’s performance. Stratified five-fold cross-validation was used to evaluate the performance of the proposed method. The mean accuracy, sensitivity, specificity and Matthews correlation coefficient observed on the PhysioNet/CinC Challenge 2016 dataset were 86.8%, 87%, 86.6% and 72.1% respectively. The proposed algorithm’s performance achieves an appropriate trade-off between sensitivity and specificity.

Highlights

  • Statistics show that cardiovascular disease (CVD) is one of the main reasons for mortality in the world [1,2]

  • The sensitivity, specificity, MCC [42] and mean accuracy (Macc) for each fold of the cross-validation were used to evaluate the performance of the proposed model

  • Because the classification task involved a class imbalance, we applied the arithmetic mean of sensitivity and specificity, that is, the mean accuracy (Macc), to eliminate any reported high virtual accuracy

Read more

Summary

Introduction

Statistics show that cardiovascular disease (CVD) is one of the main reasons for mortality in the world [1,2]. Heart sounds are a kind of mechanical vibrations caused by the movement of blood in the cardiovascular system [3] and they are considered to be an important indicator in the diagnosis of several CVDs. Compared with electrocardiograph (ECG) signals, a phonocardiogram (PCG), which is a graphical recording of heart sounds, can clearly express changes in the state of the heart. PCG can be used to diagnose deformation in heart organs and damage to heart valves [4]. The results of an auscultation are used by doctors to detect cardiac diseases. This process is subjective, time-consuming and requires

Methods
Results
Discussion
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call