Full-hexahedral structured meshing for image-based computational vascular modeling.

Area of Science:

• Biomedical Engineering
• Computational Science
• Medical Imaging

Background:

• Image-based computational modeling provides high-resolution in vivo data for flow and structural mechanics.
• Patient-specific modeling relies on accurate mesh generation, often using unstructured tetrahedral meshes.
• Structured hexahedral meshes offer superior accuracy and computational efficiency but are difficult to apply to complex vascular geometries.

Purpose of the Study:

• To develop a robust procedure for generating structured hexahedral meshes tailored to subject-specific vascular topology.
• To enable more accurate and computationally efficient Finite Element Analysis (FEA) and Computational Fluid Dynamics (CFD) in cardiovascular studies.
• To create a versatile meshing tool applicable to complex human vascular cases.

Main Methods:

• A bottom-up approach using centerline-based synthetic descriptors (centerlines, radii, normals).
• Placement of topologically equivalent block-structures and a projection operation for parametric volume mesh generation.
• Artificial generation of a three-layered arterial wall with variable thickness and material anisotropy.

Main Results:

• Successful generation of high-quality structured hexahedral meshes for challenging cases like the aortic arch and abdominal aortic aneurysm.
• Meshes achieved sub-micrometric accuracy, fitting original triangulated models precisely.
• Demonstrated improvement in shape-quality metrics for the generated meshes.

Conclusions:

• The developed procedure offers a robust solution for structured mesh generation in patient-specific vascular modeling.
• This tool enhances computational accuracy, potentially increasing confidence in vascular simulation results.
• The open-source implementation makes advanced meshing techniques more accessible for cardiovascular research.