Large-scale subject-specific cerebral arterial tree modeling using automated parametric mesh generation for blood...

Area of Science:

• Biomedical Engineering
• Medical Imaging
• Computational Fluid Dynamics

Background:

• Accurate 3D modeling of cerebral vasculature is crucial for understanding blood flow dynamics.
• Existing methods for generating computational meshes of cerebral arteries are often time-consuming and may lack anatomical precision.
• Simulating blood flow in the entire cerebral arterial tree, including smaller vessels, presents significant computational challenges.

Purpose of the Study:

• To develop and validate a novel, automated parametric meshing technique for generating subject-specific computational models of the cerebral arterial tree.
• To improve the quality and anatomical accuracy of meshes for efficient blood flow simulations.
• To extend 3D vascular modeling capabilities to encompass a larger portion of the cerebral arterial network.

Main Methods:

• A parametric meshing procedure was developed to automatically decompose vascular skeletons and extract geometric features.
• Hexahedral meshes were generated using a body-fitted coordinate system that respects vascular network topology.
• Statistical analysis, including receiver operating characteristic (ROC) curves, was used to validate anatomical accuracy against raw MRA data.

Main Results:

• The developed technique successfully generated high-quality, anatomically accurate computational meshes of subject-specific cerebral arterial trees.
• Geometric accuracy evaluation demonstrated strong agreement with raw MRA data, achieving an area under the curve (AUC) of 0.87.
• Parametric meshing resulted in significant improvements in mesh quality, with an average of 36.6% improvement in orthogonal skew and 21.7% in equiangular skew compared to unstructured tetrahedral meshes.

Conclusions:

• The automated parametric meshing pipeline enables efficient reconstruction and simulation of blood flow in large portions of the cerebral arterial tree, down to pial vessels.
• This technique represents a significant advancement towards fast, large-scale, subject-specific hemodynamic analysis for clinical applications.
• The study lays the groundwork for enhanced diagnostic and therapeutic strategies in cerebrovascular diseases.