Related Experiment Videos
Noninvasive detection of coronary stenoses before and after angioplasty using eigenvector methods
M Akay1, J L Semmlow, W Welkowitz
1Department of Biomedical Engineering, Rutgers University, Piscataway, NJ 08855.
IEEE Transactions on Bio-Medical Engineering
|November 1, 1990
Summary
Eigenvector methods revealed reduced high-frequency heart sounds after coronary angioplasty, supporting turbulent blood flow theories in partially occluded arteries. This noise-robust analysis overcomes limitations of traditional FFT methods.
Area of Science:
- Biomedical Engineering
- Cardiovascular Physiology
- Signal Processing
Background:
- Partially occluded coronary arteries may generate audible sounds from turbulent blood flow.
- Traditional Fast Fourier Transform (FFT) analysis struggles with noise in low-level heart sound signals.
- Previous studies using FFT on this data showed no significant changes post-angioplasty.
Purpose of the Study:
- To compare frequency spectra of diastolic heart sounds before and after angioplasty.
- To investigate the efficacy of advanced eigenvector methods for analyzing noisy heart sounds.
- To support the hypothesis that turbulent blood flow in occluded arteries produces distinct sound signatures.
Main Methods:
- Applied three high-resolution eigenvector methods: Pisarenko, MUSIC, and Minimum-Norm, to analyze heart sound frequency spectra.
- Compared spectral components before and after coronary angioplasty surgery.
- Excluded the Pisarenko method due to spurious zero generation.
Main Results:
- MUSIC and Minimum-Norm eigenvector methods successfully generated frequency spectra from noisy heart sound data.
- These methods revealed a significant decrease in high-frequency spectral components post-angioplasty in most cases.
- This contrasts with previous findings using traditional FFT analysis.
Conclusions:
- Eigenvector methods offer superior resolution for analyzing heart sounds in the presence of noise.
- The observed reduction in high-frequency components supports the link between turbulent flow in coronary arteries and heart sounds.
- Advanced signal processing techniques can enhance the diagnostic potential of phonocardiography.