Abstract
Hemodynamic predictions using computational fluid dynamics (CFD) simulations can provide valuable guidance assessing aortic disease risks. However, their reliability is hindered by the lack of patient-specific boundary conditions, particularly measured flow rates. This study addresses this knowledge gap by introducing a method for estimating flow division in aortic branches. The geometry of the lesional aorta was first repaired to obtain a near-healthy reference geometry. An iterative CFD simulation was then employed to estimate the flow division in the branches of the diseased aorta. Specifically, empirical boundary conditions from healthy individuals were used to predict the outlet pressures of reference geometry, which were subsequently converted into resistance models. These resistance models were then assigned to the outlets of the diseased aorta to predict the inlet pressure. The discrepancy between the predicted and target inlet pressures was iteratively minimized by adjusting the inlet pressure of the reference model until convergence was achieved. The final flow division in the branches of the diseased aorta was then obtained. The performance of the proposed method was investigated in three patients with aortic dissection or aneurysm. The proposed method predicted lower flow rates in branches with severe stenosis, which was more consistent with physiological expectations. Furthermore, the predicted blood pressure differed significantly from that obtained using the traditional method and was closer to the target values. The proposed method provides a practical solution for specifying boundary conditions in hemodynamic studies when clinically measured flow rates are unavailable.