BACKGROUND: The study was designed to test the angle independence of a dynamic three-dimensional digital colour Doppler method for laminar flow measurement. The technique acquired three-dimensional data by rotational acquisition and used surface integration of Doppler vector velocities and flow areas in time and space for flow computation. METHOD: A series of pulsatile flows (peak flow 55-180 ml/s) through a curved tube were studied with reference flow rates obtained using an ultrasonic flow meter. Colour Doppler imaging was performed at three angles to the direction of flow (20 degrees, 30 degrees, 40 degrees), using a multiplane transoesophageal probe controlled by an ATL HDI5000 system. Integration of digital velocity vectors over a curved three-dimensional surface across the tube for each of the 11 flow rates at each angle was performed off-line to compute peak flow. RESULTS: Peak flow rates correlated closely (r=0.99) with the flow meter with the mean difference from the reference being -0.8+/-2 x 4 ml/s, 0.9+/-2.6 ml/s, 1.0+/-2 x 3 ml/s for 20 degrees, 30 degrees and 40 degrees respectively. Comparison of the three angle groups showed no significant differences (P=0.15, ANOVA). When sampled obliquely, the flow area on the curved surface increased while the velocities measured decreased. CONCLUSION: Surface integration of velocity vectors to compute three-dimensional Doppler flow data is less angle dependent than conventional Doppler methods.
BACKGROUND: The study was designed to test the angle independence of a dynamic three-dimensional digital colour Doppler method for laminar flow measurement. The technique acquired three-dimensional data by rotational acquisition and used surface integration of Doppler vector velocities and flow areas in time and space for flow computation. METHOD: A series of pulsatile flows (peak flow 55-180 ml/s) through a curved tube were studied with reference flow rates obtained using an ultrasonic flow meter. Colour Doppler imaging was performed at three angles to the direction of flow (20 degrees, 30 degrees, 40 degrees), using a multiplane transoesophageal probe controlled by an ATL HDI5000 system. Integration of digital velocity vectors over a curved three-dimensional surface across the tube for each of the 11 flow rates at each angle was performed off-line to compute peak flow. RESULTS: Peak flow rates correlated closely (r=0.99) with the flow meter with the mean difference from the reference being -0.8+/-2 x 4 ml/s, 0.9+/-2.6 ml/s, 1.0+/-2 x 3 ml/s for 20 degrees, 30 degrees and 40 degrees respectively. Comparison of the three angle groups showed no significant differences (P=0.15, ANOVA). When sampled obliquely, the flow area on the curved surface increased while the velocities measured decreased. CONCLUSION: Surface integration of velocity vectors to compute three-dimensional Doppler flow data is less angle dependent than conventional Doppler methods.