Automatic Reconstruction and Generation of Structured Hexahedral Mesh for Non-planar Bifurcations in Vascular Networks

Abstract

Abstract Unstructured meshes are widely used to delineate the complex geometry of blood vessels in computational fluid dynamics simulation. However, structured parametric meshes can better represent complex geometries with lower grid density thus enabling faster and more precise computations. Current mesh generation methods require user interaction and cannot be used for automatic generation of parametric mesh representations of vascular networks. Our aim is to present a fully automatic algorithm for parametric mesh representation of vascular networks from subject-specific medical images. The methodology requires no user interaction, can create volumetric meshes for non-planar bifurcations and offers full control over the local mesh resolution at the surface and the lumen of the blood vessels. We demonstrate the reliability of our algorithm in terms of mesh quality by evaluating scaled Jacobian and equiangular skewness. The novel method was applied in case studies to reconstruct real image data for the cerebral angioarchitecture. Accurate subject-specific representations of vascular trees are necessary to perform organ-wide hemodynamics for personalized surgical planning of vascular disease.