Validation of numerical simulation methods in aortic arch using 4D Flow MRI.

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Computational Science

Background:

• Computational fluid dynamics (CFD) is crucial for blood flow dynamics but relies heavily on boundary conditions and computational models.
• Validation of CFD models against in-vivo measurements is essential for reliable simulation of hemodynamics.

Purpose of the Study:

• To validate computational fluid dynamics (CFD) methods by comparing simulation results with actual blood flow measurements.
• To assess the accuracy of different CFD turbulent models against four-dimensional (4D) flow magnetic resonance imaging (MRI) data.

Main Methods:

• 4D Flow MRI was performed on a healthy adult and a child with a double-aortic arch.
• Aortic lumen segmentation was used for CFD analysis with laminar, Large Eddy Simulation (LES), and Renormalization Group k-ε (RNG k-ε) turbulent models.
• Comparison of flow-velocity vector components, wall shear stress (WSS), and energy loss (EL) between MRI and CFD results.

Main Results:

• The RNG k-ε model showed the highest correlation coefficients for velocity vector components (up to 0.768 in the aortic arch) in the child patient.
• CFD simulations accurately predicted flow distal to the aortic arch and WSS distribution patterns, though helical flow in the ascending aorta was not captured.
• WSS and EL values from MRI were lower than CFD, but distribution patterns were similar; turbulence models (RNG k-ε, LES) yielded higher estimates than the laminar model.

Conclusions:

• CFD simulations, particularly with turbulence models like RNG k-ε, can accurately represent blood flow dynamics and WSS distribution compared to 4D Flow MRI.
• The RNG k-ε model offers enhanced accuracy for CFD validation in complex cardiovascular geometries.
• Further refinement is needed to accurately simulate complex flow patterns like helical flow in the ascending aorta using CFD.