Phonocardiogram signal analysis: techniques and performance comparison

Area of Science:

• Biomedical Engineering
• Signal Processing
• Cardiology

Background:

• Phonocardiogram (PCG) signals contain vital diagnostic information about cardiac function.
• Accurate analysis of PCG signals requires advanced time-frequency representation methods.
• Traditional methods like spectrograms and Wigner distribution have limitations in resolving subtle PCG components.

Purpose of the Study:

• To compare the efficacy of spectrogram, Wigner distribution, and wavelet transform for PCG signal analysis.
• To evaluate the ability of each method to detect key components of heart sounds (S1 and S2).
• To determine which method provides superior time-frequency characteristics for clinical diagnosis.

Main Methods:

• Application of spectrogram (Short-Time Fourier Transform - STFT), Wigner distribution, and wavelet transform to PCG signals.
• Comparative analysis of the time-frequency resolution and diagnostic capabilities of each method.
• Focus on the detection of the four components of the first heart sound (S1) and the two components of the second heart sound (S2).

Main Results:

• Spectrogram (STFT) failed to detect all four components of S1 and inaccurately detected S2 components.
• Wigner distribution provided time-frequency characteristics but inaccurately detected S1 components and merged S2 components.
• Wavelet transform successfully detected the aortic (A2) and pulmonary (P2) components of S2, which were undetectable by other methods.
• Standard Fourier transform identified S2 components in frequency but not their temporal delay.

Conclusions:

• Wavelet transform offers superior resolution and diagnostic capability for PCG signal analysis compared to spectrograms and Wigner distribution.
• The ability of wavelet transform to detect individual S2 components (A2, P2) is crucial for accurate cardiac assessment.
• Wavelet transform provides richer time-frequency features, enabling enhanced qualitative and quantitative measurements for physicians.