Estimation of inlet flow rates for image-based aneurysm CFD models: where and how to begin?

Patient-specific flow rates are rarely available for image-based computational fluid dynamics models. Instead, flow rates are often assumed to scale according to the diameters of the arteries of interest. Our goal was to determine how choice of inlet location and scaling law affect such model-based estimation of inflow rates. We focused on 37 internal carotid artery (ICA) aneurysm cases from the Aneurisk cohort. An average ICA flow rate of 245 mL min(-1) was assumed from the literature, and then rescaled for each case according to its inlet diameter squared (assuming a fixed velocity) or cubed (assuming a fixed wall shear stress). Scaling was based on diameters measured at various consistent anatomical locations along the models. Choice of location introduced a modest 17% average uncertainty in model-based flow rate, but within individual cases estimated flow rates could vary by >100 mL min(-1). A square law was found to be more consistent with physiological flow rates than a cube law. Although impact of parent artery truncation on downstream flow patterns is well studied, our study highlights a more insidious and potentially equal impact of truncation site and scaling law on the uncertainty of assumed inlet flow rates and thus, potentially, downstream flow patterns.

Fontan hemodynamics: importance of pulmonary artery diameter.

OBJECTIVE: We quantify the geometric and hemodynamic characteristics of extracardiac and lateral tunnel Fontan surgical options and correlate certain anatomic characteristics with their hemodynamic efficiency and patient cardiac index. METHODS AND RESULTS: The study was conducted retrospectively on 22 patients undergoing Fontan operations (11 extracardiac and 11 lateral tunnel operations). Total cavopulmonary connection geometric parameters such as vessel areas, curvature, and offsets were quantified using a skeletonization method. Energy loss at the total cavopulmonary connection junction was available from previous in vitro experiments and computational fluid dynamic simulations for 5 and 9 patients, respectively. Cardiac index data were available for all patients. There was no significant difference in the mean and minimum cross-sectional vessel areas of the pulmonary artery between the extracardiac and lateral tunnel groups. The indexed energy dissipation within the total cavopulmonary connection was strongly correlated to minimum cross-sectional area of the pulmonary arteries (R(2) value of 0.90 and P < .0002), whereas all other geometric features, including shape characteristics, had no significant correlation. Finally, cardiac index significantly correlated with the minimum pulmonary artery area (P = .006), suggesting that total cavopulmonary connection energy losses significantly affect resting cardiac output. CONCLUSIONS: The minimum outlet size of the total cavopulmonary connection (ie, minimum cross section of pulmonary artery) governs the energy loss characteristics of the total cavopulmonary connection more strongly than variations in the shapes corresponding to extracardiac and lateral tunnel configurations. Differences in pulmonary artery sizes must be accounted for when comparing energy losses between extracardiac and lateral tunnel geometries.

Quantitative analysis of extracardiac versus intraatrial Fontan anatomic geometries.

BACKGROUND: There exists large geometric variability among total cavopulmonary connections (TCPC) because of the patient-specific anatomies and the chosen surgical procedure. In this study we present quantitative comparison of the geometric characteristics of the extracardiac and intraatrial Fontan anatomies, the two commonly used TCPC procedures. METHODS: A method of centerline approximation of the three-dimensional geometries (skeletonization) was used to quantify the TCPC geometric parameters such as vessel areas, curvature, and collinearity. The TCPC anatomies of 26 patients, 13 extracardiac and 13 intraatrial, were analyzed in this study. RESULTS: There was no significant difference in the vessel dimensions between extracardiac and intraatrial TCPCs, with the overall magnitudes agreeing well with that seen in normal children except for the inferior vena cava. Intraatrial baffles had significant fluctuations in cross-sectional area along the length of the baffle as opposed to extracardiacs (p < 0.05). Patients with hypoplastic left heart syndrome had significant narrowing of the left pulmonary artery (p < 0.05), suggesting a possible physical constriction from the reconstructed aorta. CONCLUSIONS: This study benchmarks the anatomic variability of patient-specific TCPCs. Intraatrial Fontan geometries have significant difference in the area variations across the vessel length compared with the extracardiac geometry. Also, patients with hypoplastic left heart are at a higher risk of left pulmonary artery narrowing.