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

Computational Fluid Dynamics (CFD) is widely used in both scientific and industrial contexts to provide valuable insights into cardiovascular flows. CFD supports the development of medical devices, describes complex flow physical phenomena associated with disease generation and progression, and even informs surgical planning. Non-invasive technologies such as 4D MRI provide detailed information about blood flow for a given patient, yet CFD allows higher spatial and temporal resolution less invasively. However, the advantages of CFD methods can only be realized through faithful geometry reconstruction, high-quality mesh generation, and suitable definition of patient- specific inlet and outlet boundary conditions. In this regard, 4D MRI measurements can provide the required data to calibrate and validate patient-specific CFD models. Hence, the combination of 4D MRI and CFD is crucial for accurate and efficient in-silico cardiovascular flow predictions for patient-specific geometries. Multiple CFD software, such as ANSYS, COMSOL, and SimVascular, have been widely used in published research works, yet the graphical user interfaces of these software packages have no explicit provisions for leveraging spatio-temporal 4D MRI data. Here, we present a framework to create accurate patient-specific CFD simulations leveraging spatiotemporal 4D MRI data and patient monitoring data. We discuss all aspects of patient-specific modeling, including geometry reconstruction, meshing, numerical simulation, and post-processing of results, focusing on the suitability of each software package and the ease with which the presented workflow can be implemented with the software.