Session: 05-01-02 Applied CFD

Paper Number: 87596

87596 - Combining 4D MRI With CFD for Investigating Patient-Specific Cardiovascular Flows: A Comprehensive Comparison of ANSYS, COMSOL, and SimVascular Illustrated With the Prediction of Thoracic Aortic Hemodynamics 

In spite of the advancement of medical imaging, particularly four-dimensional magnetic resonance imaging (4D MRI), computational fluid dynamics (CFD) is still considered the main research approach for numerical (in-silico) modeling to provide valuable insights into cardiovascular flows, such as in thoracic aorta. It is owing to the fact that CFD is capable of providing a higher spatial and temporal resolution in comparison with 4D MRI. However, this higher CFD resolution can only be achieved by appropriate geometry reconstruction, high-quality mesh generation, and proper selection of inlet and outlet boundary conditions for which 4D MRI can provide the required data to calibrate and verify the CFD patient-specific models. As a result, combining 4D MRI and CFD seems crucial for accurate and efficient in-silico flow predictions for patient-specific geometries. In this regard, multiple CFD software such as ANSYS, COMSOL, and SimVascular, which are based on either finite volume method (FVM) or finite element method (FEM), have been widely utilized for investigating cardiovascular flows and their corresponding hemodynamic indices. However, these CFD solvers do not provide a graphical user interface option for directly coupling with spatiotemporal 4D MRI data.   Therefore, the overall objective of this paper is to provide a framework explaining how spatiotemporal 4D MRI data, such as inlet boundary conditions, can be incorporated with some of the well-known CFD packages (ANSYS, COMSOL, and SimVascular) for patient-specific simulations. This paper also describes other aspects of the CFD modeling, including patient-specific geometry reconstruction, meshing, numerical scheme, flexibility in controlling the numerical schemes, ease of parallelization, preprocessing, simulation and postprocessing, and estimated computational cost. Ultimately, researchers in the field of cardiovascular flows who wish to combine 4D MRI with CFD can opt for their most suitable CFD package after reviewing this comprehensive comparison.

Presenting Author: Farshad Tajeddini University of Toronto