Improvement of the Accuracy in the Identification of Coronary Artery Disease Combining Heart Sound Features

Area of Science:

• Cardiology
• Biomedical Engineering
• Signal Processing

Background:

• Diastolic murmurs are commonly used for coronary artery disease (CAD) diagnosis.
• However, CAD diastolic murmurs are often weak and susceptible to noise and valvular murmurs, limiting diagnostic accuracy.
• Existing methods solely relying on diastolic murmurs show suboptimal performance in CAD identification.

Purpose of the Study:

• To develop and validate an enhanced algorithm for improved coronary artery disease diagnosis.
• To combine features from the first heart sound and diastolic murmurs for more accurate CAD identification.
• To overcome the limitations of using only diastolic murmurs in diagnosing coronary artery disease.

Main Methods:

• A first heart sound feature extraction algorithm was employed for initial CAD identification.
• Empirical Wavelet Transform decomposed diastolic heart sounds into three modes for spectral energy analysis.
• Features from the first heart sound, second diastolic spectral energy, and a parameter (P3) for murmur discrimination were integrated.

Main Results:

• The proposed algorithm demonstrated improved accuracy in identifying coronary artery disease.
• Integration of first heart sound and specific diastolic murmur features enhanced diagnostic performance.
• Experimental comparisons showed the algorithm outperformed several state-of-the-art methods.

Conclusions:

• Combining first heart sound and diastolic murmur features significantly improves coronary artery disease diagnostic accuracy.
• The proposed algorithm offers a more robust and accurate approach to CAD diagnosis compared to traditional methods.
• This approach holds promise for enhancing non-invasive detection of coronary artery disease.