Hemodynamic evaluation of arterial stenoses by computer simulation.

Abstract

A new method to assess the hemodynamic severity of arterial stenoses was proposed and evaluated. It is based on a previously developed finite element computer simulation model for laminar-separated flow in arteries of axially varying cross-section; the present modification allows use of angiographic stenosis shapes acquired by automatic edge-detection algorithms. The method was validated by comparing its results with published experimental and theoretic results for ideal stenosis shapes. At moderate flowrates (Reynolds number = 500), poststenosis flow separation was predicted for moderately severe (75% area reduction) but not for mild (25%) stenoses. For high flowrates (Reynolds number = 900) in a severe stenosis (89%), stagnation and reversed flow were predicted and the experimental nondimensional pressure drop of 48.5 was correctly determined. Bernoulli's Equation, which neglects viscosity, predicted a drop of only 40. For a severe stenosis (89%), even at low Reynolds numbers (50), reversed flow agreeing with other theoretic solutions was predicted. Predictions are especially useful at low flow rates, where experiments are difficult to conduct. The height of the curve on the graph of nondimensional pressure gradient vs. Reynolds number reflects the hemodynamic severity of a particular stenosis; these curves were predicted for moderate and severe ideal stenoses and agree with experiments. A similar analysis is applied to an actual human coronary artery stenosis, and the results are demonstrated to have use in assessing interventions during angiography.