Automatic Diagnosis of Cardiovascular Diseases Through Analysis of Heart Sound Signals Using CNN Models: A Survey (Special Issue: 3rd Young Researchers Conference 2025 ) | ||||
| International Journal of Applied Energy Systems | ||||
| Volume 8, Issue 1, January 2026, Page 15-20 PDF (2.25 MB) | ||||
| Document Type: Original papers | ||||
| DOI: 10.21608/ijaes.2026.442097 | ||||
 View on SCiNiTO | ||||
| Authors | ||||
Aya Bakr  ; Samia Heshmat
| ||||
| Department of Electrical Engineering Faculty of Engineering, Aswan University Aswan, Egypt | ||||
| Abstract | ||||
| Heart diseases often cause changes in heart sounds and murmurs before other symptoms appear, making auscultation a crucial first step in diagnosing cardiovascular conditions. However, heart sound analysis has not been widely adopted due to uncertainties about the origins of these sounds and the lack of reliable quantitative methods for analyzing them. Since heart sound signals contain much more information than the human ear or traditional visual inspection methods can detect, automated classification is essential for early detection, especially in primary healthcare settings. This paper explores the use of deep Convolutional Neural Networks (CNNs) for classifying heart sounds as normal or abnormal. It provides a detailed analysis of CNN-based approaches, highlighting their strengths in feature extraction and classification accuracy compared to conventional methods. The paper also discusses key challenges, including model generalization, data quality, and integration with other diagnostic tools. By reviewing recent advancements, this study emphasizes the potential of CNNs to improve early diagnosis and enhance patient outcomes in cardiovascular health. | ||||
|
Statistics Article View: 70 PDF Download: 36 |
||||
