[A heart sound classification method based on complete ensemble empirical modal decomposition with adaptive noise...

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Heart sound signals (phonocardiograms, PCG) are complex physiological signals.
• These signals exhibit nonlinear and nonstationary characteristics, posing challenges for accurate classification.
• Existing methods for PCG analysis may lack efficiency and precision.

Purpose of the Study:

• To develop a novel and efficient method for phonocardiogram (PCG) classification.
• To improve the accuracy of heart sound signal analysis.
• To leverage advanced signal decomposition and entropy analysis for enhanced feature extraction.

Main Methods:

• Utilized Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) to decompose PCG signals into intrinsic mode functions (IMFs).
• Sifted IMFs based on correlation coefficient, energy factor, and signal-to-noise ratio for optimal selection.
• Extracted instantaneous frequency via Hilbert transform and computed permutation entropy to form eigenvectors for Support Vector Machine (SVM) classification.

Main Results:

• The proposed CEEMDAN permutation entropy method achieved an accuracy rate of up to 87% in PCG classification.
• Demonstrated a significant improvement in accuracy, an increase of 18%-24%, compared to traditional Empirical Mode Decomposition (EMD) and Ensemble Empirical Mode Decomposition (EEMD) methods.
• Validated the method's effectiveness using 100 PCG samples from the 2016 PhysioNet/CinC Challenge.

Conclusions:

• The CEEMDAN permutation entropy method offers a highly accurate and efficient approach for phonocardiogram classification.
• This technique effectively captures the nonlinear and nonstationary features of heart sound signals.
• The proposed method represents a significant advancement in automated cardiac auscultation and diagnosis.