Related Experiment Video
Updated: Oct 10, 2025
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Segmentation-free Heart Pathology Detection Using Deep Learning
Insights
This study introduces a novel, segmentation-free method for classifying heart sounds, improving automated cardiovascular diagnosis. The approach enhances precision for normal and murmur heart sounds, showing potential for practical clinical applications.
Area of Science:
- Cardiology
- Biomedical Engineering
- Signal Processing
Background:
- Cardiovascular diseases are a leading global cause of mortality.
- Heart sound auscultation is crucial for cardiovascular examination but difficult to master.
- Existing automated methods often fail with noisy signals or high heart rates due to reliance on segmentation.
Purpose of the Study:
- To develop a novel segmentation-free heart sound classification method.
- To improve the accuracy and robustness of automated cardiovascular diagnosis.
- To enable practical, automatic detection of heart murmurs.
Main Methods:
- Applied discrete wavelet transform for signal denoising.
- Performed feature extraction and reduction.
- Utilized Support Vector Machines and Deep Neural Networks for classification.
Main Results:
- Achieved 81% precision for normal and 96% for murmur classes on the PASCAL heart sound dataset.
- Demonstrated superior performance compared to existing methods.
- Achieved 92% precision for normal and 86% for murmur in a user-independent setting.
Conclusions:
- The proposed segmentation-free method offers superior performance in heart sound classification.
- The approach shows significant potential for practical, automatic murmur detection.
- This method addresses limitations of previous techniques, especially in challenging signal conditions.
Abstract:
Cardiovascular (CV) diseases are the leading cause of death in the world, and auscultation is typically an essential part of a cardiovascular examination. The ability to diagnose a patient based on their heart sounds is a rather difficult skill to master. Thus, many approaches for automated heart auscultation have been explored. However, most of the previously proposed methods involve a segmentation step, the performance of which drops significantly for high pulse rates or noisy signals. In this work, we propose a novel segmentation-free heart sound classification method. Specifically, we apply discrete wavelet transform to denoise the signal, followed by feature extraction and feature reduction. Then, Support Vector Machines and Deep Neural Networks are utilised for classification. On the PASCAL heart sound dataset our approach showed superior performance compared to others, achieving 81% and 96% precision on normal and murmur classes, respectively. In addition, for the first time, the data were further explored under a user-independent setting, where the proposed method achieved 92% and 86% precision on normal and murmur, demonstrating the potential of enabling automatic murmur detection for practical use.
