4D Flow MRI-enabled patient-specific computational hemodynamics of thoracic aorta: CFD predictions vs. in vivo imaging data

Abstract

The assessment of hemodynamic parameters within the human vasculature is crucial for understanding cardiovascular health and disease progression. This study employs patient-specific 4D Flow MRI data to validate computational fluid dynamics (CFD) simulations of blood flow in the thoracic aorta. Two CFD solvers, SimVascular and CRIMSON, are utilized to simulate hemodynamics using both patient-specific and parabolic inlet velocity profiles. The objective is to evaluate the accuracy of computational predictions in comparison to in vivo imaging data. The methodology includes 3D reconstruction of the thoracic aorta from MRI data, preprocessing of velocity fields, and numerical simulations incorporating appropriate boundary conditions. A direct comparison between simulated and MRI-derived velocity fields, pressure distributions, and wall shear stress (WSS) is conducted at multiple time points throughout the cardiac cycle. Special emphasis is placed on analyzing flow characteristics in critical regions, including the ascending aorta, aortic arch, and descending aorta. Results indicate that while both CFD solvers accurately capture global flow trends, differences arise in local flow patterns, particularly in regions with complex hemodynamics. The patient-specific velocity profile demonstrates superior agreement with MRI data, especially in replicating secondary flow structures and WSS distributions. Conversely, the parabolic profile tends to overestimate peak velocities and introduce deviations in flow patterns. The analysis also highlights minor discrepancies between CRIMSON and SimVascular, likely attributed to differences in mesh resolution and numerical techniques. The findings emphasize the importance of patient-specific boundary conditions in cardiovascular simulations, as they significantly impact flow field accuracy. The study underscores the strengths and limitations of CFD modeling for clinical applications and highlights the potential of 4D Flow MRI as a validation tool for computational hemodynamics. Future work should explore the integration of fluid-structure interaction (FSI) models to enhance simulation realism by incorporating vessel wall compliance.