High-Fidelity Tetrahedral Mesh Generation from Medical Imaging Data for Fluid-Structure Interaction Analysis of Cerebral Aneurysms

This paper describes a comprehensive and high-fidelity finite element meshing approach for patient-specific arterial geometries from medical imaging data, with emphasis on cerebral aneurysm configurations. The meshes contain both the blood volume and solid arterial wall, and are compatible at the fluid-solid interface. There are four main stages for this meshing method: 1) Image segmentation and geometric model construction; 2) Tetrahedral mesh generation for the fluid volume using the octree-based method; 3) Mesh quality improvement stage, in which edge-contraction, pillowing, optimization, geometric flow smoothing, and mesh cutting are applied to the fluid mesh; and 4) Mesh generation for the blood vessel wall based on the boundary layer generation technique. The constructed meshes are extensively employed in a fully-coupled fluid-structure interaction analysis of vascular blood flow. This paper presents several case studies of hemodynamics in patient-specific cerebral aneurysms.