Carotid artery volume flow: in vivo measurement with time-domain-processing US

To evaluate carotid artery volume-flow measurements with time-domain-processing ultrasonography (US).

Volume-flow measurements were obtained in the carotid arteries of nine swine with time-domain-processing US. Four swine were prepared with a model of an arteriovenous shunt. Flow through the carotid artery was varied by means of a number of physiologic and pharmacologic interventions. The reference standard for volume flow consisted of measurements with transit-time US flowmetry. At a limited number of measurement points, this reference standard was validated against true volume flow measured with timed collection.

For flow rates less than 500 mL/min, a linear correlation existed between measurements with time-domain processing and the US flow meter (r2 = .96, P < .001, slope = 1.117). Values above 500 mL/min were less well correlated with the reference standard. True flow was underestimated with both methods, less so with time-domain processing than with US flowmetry (underestimation, 10% versus 21%). Measurements were significantly less reproducible with time-domain processing than with US flowmetry (P < .001). Interobserver variability was negligible.

Because of operator errors, measurements with time-domain processing should be repeated at least three times to ensure accuracy and may be inaccurate in flow rates over 500 mL/min.