Hemodynamically significant coronary artery stenoses generate turbulent blood flow patterns that manifest as intracoronary murmurs. This study aims to evaluate the performance of modern acoustic detection of these murmurs by acoustic signals captured from patients undergoing gold standard comparative coronary angiography.

We prospectively studied 156 patients undergoing elective coronary angiography, excluding those with acute coronary syndrome, prior chest surgery, or significant valvular disease. Acoustic signals were captured before arterial access. Angiographic degree of stenosis in each coronary artery was graded blinded to clinical and acoustic data. Acoustic data were analyzed blinded to clinical and angiographic data, categorizing subjects as “normal,” “diseased,” or “inconclusive.” Of 156 patients examined, 123 generated analyzable data.

Angiographically significant stenosis (≥50%) prevalence was 52% (18%, 23%, 11% with 1-, 2-, 3-vessel disease, respectively). Acoustic detection sensitivity and specificity for stenosis ≥50% in any vessel were 0.70 and 0.80, respectively (negative predictive value, 0.71; positive predictive value, 0.79). Acoustic detection optimally identified stenosis ≥50% with an area under the curve of 0.75. For stenosis ≥50% in major vessels only (left main, proximal-mid left anterior descending, proximal-mid circumflex, proximal-mid right coronary), prevalence was 46%; sensitivity and specificity were 0.72 and 0.76, respectively (negative predictive value, 0.76; positive predictive value, 0.72; area under the curve, 0.76).

Acoustic signal patterns and modern analysis techniques may be used to identify intracoronary murmurs generated by hemodynamically significant coronary artery stenoses in all major vessels. Further investigation is warranted to compare the clinical performance of this modality with current noninvasive approaches that evaluate patients at risk for atherosclerotic and obstructive coronary artery disease.