Comparison of Doppler echocardiographic peak frequency and turbulence parameters in the quantification of aortic stenosis in a pulsatile flow model.

Abstract

To test the relative accuracy of Doppler echocardiographic peak frequency and turbulence parameters in assessing aortic stenosis, we constructed a pulsatile flow model that simulated human left ventricular and aortic pressures, flow, and anatomy. Continuous wave-measured peak frequencies and pulsed Doppler-measured turbulence were determined in the model ascending aorta for nine stenotic valve areas for each of five different flow rates. The mean squared systolic peak frequency (MSPF) and turbulence spectral envelope area (SEA) were regressed against the mean systolic gradient (r = .94, SEE = 5.6 mm Hg; and r = .96, SEE = 1.2 mm Hg, respectively). SEA was more accurate than MSPF at moderate-to-high degrees of stenosis and exhibited a smaller standard error. MSPF was more accurate than SEA in mild stenoses, where SEA tended to overestimate gradients. When flow data were included in a multiple regression analysis, both MSPF and SEA provided fair predictions of actual effective valve areas (r = .90 and r = .94, respectively). Use of high pulse-repetition-frequency Doppler echocardiography significantly reduced aliasing problems common to pulsed Doppler techniques.