An experimental evaluation of the use of an ensemble average for the calculation of turbulence in pulsatile flow.

A comparison was made between turbulence calculated by subtracting an ensemble average from the instantaneous velocity and calculations made with a high pass digital filter. Velocity was measured with a laser Doppler anemometer in vitro in the region of a normal porcine aortic valve and in patients with a hot film anemometer in the region of normal aortic valves. From the velocity obtained in patients, the absolute turbulence intensity calculated using an ensemble average of 50 beats was nearly twice the turbulence intensity calculated using a digital filter. Individual beats sometimes showed differences of 150% compared to calculations based upon the use of a digital filter. Inspection showed that the ensemble average varied widely from the actual nonfluctuating velocity. Studies in vitro showed less beat to beat variation than occurred in patients. The absolute turbulence intensity measured in vitro, when calculated using an ensemble average, was only 20% greater than calculations using a digital filter. The differences were due primarily to beat-to-beat variations of the nonfluctuating velocity, but these beat-to-beat variations were less prominent than occurred in patients. These observations suggest that ensemble averaging may not be appropriate for the calculation of turbulence, particularly in patients.