Spatial velocity distribution and acceleration in serial subvalve tunnel and valvular obstructions: An in vitro study using doppler color flow mapping

To evaluate the spatial distribution of flow velocities, turbulence and spatial acceleration in serial tunnel-valve obstruction, Doppler color flow mapping was performed in a pulsatile flow model with a tunnel obstruction (1.0 or 1.5 cm²) inserted at 2, 20 and 40 mm proximal to a mildly stenotic bioprosthetic valve studied at flow rates of 1, 2.7 and 4.9 liters/min. Measured pressure gradients were consistently higher across the tunnel (mean ± SD 32.7 ± 26.5 mm Hg) than across the tunnel plus valve (28.8 ± 26.9 mm Hg, p < 0.01). Doppler color flow map images were analyzed using a Sony RGB video-digitizing computer, providing numerical velocity assignments for the blue, red and green (variance) pixel components to allow the flow maps to be constructed into digital velocity maps and pseudo three-dimensional velocity maps. The maximal velocity stream extended distal to the tunnel (2 to 19 mm), and the length of this extension correlated well with the pressure gradient measured across the tunnel (r = 0.89), with a rapidly decelerating and turbulent spray area seen immediately distal to the valve. Pressure gradient calculated from the maximal velocity derived from the color flow map, which could only be estimated from the velocity maps for the 1.5 cm² tunnel, correlated well with the gradient measured across the tunnel (18.0 ± 14.1 versus 19.2 ± 14.5 mm Hg, respectively, r = 0.98). Acceleration was seen proximal to both tunnels. The length of the proximal acceleration (4 to 47 mm) increased with higher cardiac output (9.7 ± 4.5 mm at 1 liter/min versus 34.8 ± 8.7 mm at 4.9 liters/min, p < 0.05) and was associated with green (variance) flow on digital computer analysis. Because this area of proximal acceleration showed only laminar flow on flow visualization, to clarify this latter observation, a constant flow model (5 to 30 liters/min) with a tapering funnel (6.5°) leading into a constant velocity cylinder was also studied. The funnel produced constant spatial acceleration without the temporal acceleration that occurs with pulsatile flow. Green flow or variance occurred with the spatial acceleration, but not with the constant velocity segment despite identical maximal velocities. Hence, variance associated with proximal acceleration is not hydrodynamic turbulence, but results from a Doppler color flow mapping sampling error. Doppler color flow mapping provides valuable velocity, acceleration and turbulence observations in vitro, which may allow evaluation of serial obstructions because they do not follow the simplified Bernoulli relation.