Department of Biomedical Engineering, Erasmus Medical Center, Rotterdam 3000CA, The Netherlands.
Department of Biomedical Engineering, Erasmus Medical Center, Rotterdam 3000CA, The Netherlands.
J Biomech. 2014 Jun 3;47(8):1810-5. doi: 10.1016/j.jbiomech.2014.03.028. Epub 2014 Mar 27.
Pressure drop (△p) estimations in human coronary arteries have several important applications, including determination of appropriate boundary conditions for CFD and estimation of fractional flow reserve (FFR). In this study a △p prediction was made based on geometrical features derived from patient-specific imaging data. Twenty-two mildly diseased human coronary arteries were imaged with computed tomography and intravascular ultrasound. Each artery was modelled in three consecutive steps: from straight to tapered, to stenosed, to curved model. CFD was performed to compute the additional △p in each model under steady flow for a wide range of Reynolds numbers. The correlations between the added geometrical complexity and additional △p were used to compute a predicted △p. This predicted △p based on geometry was compared to CFD results. The mean △p calculated with CFD was 855±666Pa. Tapering and curvature added significantly to the total △p, accounting for 31.4±19.0% and 18.0±10.9% respectively at Re=250. Using tapering angle, maximum area stenosis and angularity of the centerline, we were able to generate a good estimate for the predicted △p with a low mean but high standard deviation: average error of 41.1±287.8Pa at Re=250. Furthermore, the predicted △p was used to accurately estimate FFR (r=0.93). The effect of the geometric features was determined and the pressure drop in mildly diseased human coronary arteries was predicted quickly based solely on geometry. This pressure drop estimation could serve as a boundary condition in CFD to model the impact of distal epicardial vessels.
在人体冠状动脉中，压降（△p）的估计有几个重要的应用，包括确定 CFD 的适当边界条件和估计血流储备分数（FFR）。在这项研究中，基于从患者特定的成像数据中得出的几何特征，进行了△p 预测。使用计算机断层扫描和血管内超声对 22 根轻度病变的人体冠状动脉进行成像。对每根动脉进行了三个连续的建模步骤：从直线到锥形，到狭窄，再到弯曲模型。在稳态流动下，对每个模型在广泛的雷诺数范围内进行 CFD 计算，以计算附加△p。将附加几何复杂性与附加△p 之间的相关性用于计算预测的△p。基于几何的预测△p 与 CFD 结果进行了比较。CFD 计算出的平均△p 为 855±666Pa。锥形和弯曲使总△p显著增加，在 Re=250 时，分别占总△p 的 31.4±19.0%和 18.0±10.9%。利用锥形角、最大面积狭窄和中心线的角度，我们能够在 Re=250 时生成一个很好的预测△p 估计值，平均误差为 41.1±287.8Pa，但标准偏差很高。此外，预测的△p 被用于准确估计 FFR（r=0.93）。确定了几何特征的影响，并仅基于几何形状快速预测了轻度病变的人体冠状动脉中的压降。这种压降估计可以作为 CFD 中的边界条件，用于模拟心外膜远侧血管的影响。
