Related Experiment Videos

Noninvasive acoustical detection of coronary artery disease using the adaptive line enhancer method

M Akay1, W Welkowitz, J L Semmlow

  • 1Biomedical Engineering Department, Rutgers State University, Piscataway, NJ 08855.

Insights

This study introduces a novel signal processing method to detect coronary artery disease using heart sounds. The technique accurately identified diseased patients by analyzing diastolic heart sounds, confirming high-frequency energy is linked to coronary stenosis.

Area of Science:

  • Cardiology
  • Biomedical Engineering
  • Signal Processing

Background:

  • Heart sounds may indicate occluded coronary arteries.
  • Previous analysis of diastolic heart sounds revealed additional frequency components in coronary artery disease (CAD) patients.

Purpose of the Study:

  • To explore advanced signal processing for noninvasive detection of CAD.
  • To present a new signal-processing approach using adaptive line enhancing (ALE) and spectral estimation of diastolic heart sounds.

Main Methods:

  • Utilized a two-stage process: ALE for noise reduction and spectral estimation (AR or ARMA) for parameter extraction.
  • Analyzed diastolic heart sounds recorded at the bedside.
  • Employed model parameters, including power spectral density (PSD) and AR/ARMA poles, for diagnosis.

Main Results:

  • The new method correctly identified normal and abnormal recordings in 39 out of 43 cases.
  • Confirmed an association between high-frequency energy (above 400 Hz) and coronary stenosis.

Conclusions:

  • The presented signal-processing approach shows promise for noninvasive CAD detection.
  • Diastolic heart sound analysis, particularly high-frequency components, is valuable for diagnosing coronary artery stenosis.

Related Concept Videos