High-Quality Consistent Meshing of Multi-label Datasets

Pons, Jean-Philippe; Ségonne, Florent; Boissonnat, Jean Daniel; Rineau, Laurent; Yvinec, Mariette; Keriven, Renaud

doi:10.1007/978-3-540-73273-0_17

Cited by 55 publications

(37 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The octree techniques [3,4] generate meshes with consistent multi-material junctions, however the quality of the elements at the interfaces turns out to be very poor and needs further improvements. There are methods relying on Delaunay refinement [5,6,7]. In practice, however, even though the quality of the final mesh is good, the representation of the interfaces is not (jiggles).…”

Section: Related Workmentioning

confidence: 98%

Unstructured mesh generation from the Virtual Family models for whole body biomedical simulations

Szczerba

Neufeld

Zefferer

et al. 2010

Procedia Computer Science

View full text Add to dashboard Cite

Physiological systems are inherently complex, involving multi-physics phenomena at a multitude of spatial and temporal scales. To realistically simulate their functions, detailed high quality multi-resolution often patient specific human models are required. Mesh generation has remained a central topic in finite element analysis (FEA) for a few decades now. Recent developments in high performance computing (HPC) driven by the need for multi-physics multiscale simulations of physiological systems define new challenges in this area. Even though many algorithms have been developed over years and are offered as commercial packages, they are often limited to mechanical engineering applications only. Mesh generation for human anatomical domains requires more effective and flexible techniques to tackle their greater geometrical and topological complexities. We present, evaluate and discuss several methods to generate unstructured body fitted multi-domain adaptive meshes with geometrically and topologically compatible interfaces from the segmented cross-sections of the Virtual Family models for the purpose of large scale whole body simulations. We found that an automated solution is difficult to achieve with real-image qualities, but if optimal methods are selected, good results can be achieved with minimal user-interactions. Therefore we believe that our observations can serve as guidance when choosing an optimal method for a specific application.

show abstract

Section: Related Workmentioning

confidence: 98%

Unstructured mesh generation from the Virtual Family models for whole body biomedical simulations

Szczerba

Neufeld

Zefferer

et al. 2010

Procedia Computer Science

View full text Add to dashboard Cite

show abstract

“…In [13], a lookup table is used in the case of three materials, and in the general case, triangle removal algorithms are used, allowing meshes with boundaries and holes to be created. -Delaunay refinement: These methods are an extension of typical Delaunay refinement-based algorithms to the case of multiple surfaces [14][15][16]. A mesh is iteratively created by adding vertices and updating mesh topology until a quality criteria is reached that terminates the algorithm.…”

Section: Related Workmentioning

confidence: 99%

An algorithm to mesh interconnected surfaces via the Voronoi interface

Saye

2013

Engineering with Computers

View full text Add to dashboard Cite

Many scientific and engineering problems involve interconnected surfaces meeting at junctions. For example, understanding the dynamics of a soap bubble foam can require modelling the fluid mechanics of liquid inside an intricate network of thin-film membranes. If a mesh of these surfaces is needed, the use of standard meshing algorithms often leads to voids, overlapping elements, or other artefacts near the junctions. Here, we present an algorithm to generate high-quality triangulated meshes of a set of interconnected surfaces with high surface accuracy. By capitalising on mathematical aspects of a geometric construction known as the ''Voronoi interface'', the algorithm first creates a topologically consistent mesh automatically, without making heuristic or complex decisions about surface topology. In particular, elements that meet at a junction do so by sharing a common edge, leading to simplifications in finite element calculations. In the second stage of the algorithm, mesh quality is improved by applying a short sequence of force-based smoothing, projection, and edge-flipping steps. Efficiency is further enhanced by using a locally adaptive time stepping scheme that prevents inversion of mesh elements, and we also comment on how the algorithm can be parallelised. Results are shown using a variety of examples arising from multiphase curvature flow, geometrically defined objects, surface reconstruction from volumetric point clouds, and a simulation of the multiscale dynamics of a cluster of soap bubbles. In this last example, generating high-quality meshes of evolving interconnected surfaces is crucial in determining thin-film liquid dynamics via finite element methods.

show abstract

“…The refinement process is controlled by highly customizable quality and size criteria on triangular facets and on tetrahedra. Pons et al [7] have adapted this mesh generation strategy to labelled images and have shown that the sizing criteria can be tissue-dependent leading to high-quality volume meshes of the different materials. However, in this work, 1-and 0-junctions are not explicitly handled and they are not accurately represented in the output mesh.…”

Section: Related Workmentioning

confidence: 99%

“…In this paper, we propose a feature (edge and corner) preserving extension of the Delaunay refinement algorithm introduced by Pons et al [7]. The basic idea is to explicitly sample 1-and 0-junctions before launching the Delaunay refinement and to constrain the latter to preserve these features.…”

Section: Related Workmentioning

confidence: 99%

Mesh Generation from 3D Multi-material Images

Boltcheva

Yvinec

Boissonnat

2009

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Abstract. The problem of generating realistic computer models of objects represented by 3D segmented images is important in many biomedical applications. Labelled 3D images impose particular challenges for meshing algorithms because multi-material junctions form features such as surface pacthes, edges and corners which need to be preserved into the output mesh. In this paper, we propose a feature preserving Delaunay refinement algorithm which can be used to generate high-quality tetrahedral meshes from segmented images. The idea is to explicitly sample corners and edges from the input image and to constrain the Delaunay refinement algorithm to preserve these features in addition to the surface patches. Our experimental results on segmented medical images have shown that, within a few seconds, the algorithm outputs a tetrahedral mesh in which each material is represented as a consistent submesh without gaps and overlaps. The optimization property of the Delaunay triangulation makes these meshes suitable for the purpose of realistic visualization or finite element simulations.

show abstract

High-Quality Consistent Meshing of Multi-label Datasets

Cited by 55 publications

References 19 publications

Unstructured mesh generation from the Virtual Family models for whole body biomedical simulations

Unstructured mesh generation from the Virtual Family models for whole body biomedical simulations

An algorithm to mesh interconnected surfaces via the Voronoi interface

Mesh Generation from 3D Multi-material Images

Contact Info

Product

Resources

About