Font Size: a A A

3D Mesh Generation Of Two-phase Bone From CT Data

Posted on:2014-08-03Degree:MasterType:Thesis
Country:ChinaCandidate:T A LiuFull Text:PDF
GTID:2268330425981055Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
Skeletal mesh generation not only has vital significance but also need to be addressed inthe human three-dimensional reconstruction. With the increase of bone damage probability(all kinds of car accident, osteoma, etc.), some bones (such as pelvic) with complex shape andanatomical structure, to better understand stress distribution and damage mechanism of boneunder different external force, and to provide an accurate and reliable model and analyticaltool for the research of bone injury, mesh generation of is necessary. At now, mesh generationfor bone is still in its infancy. By the thinking of doctor from CT image can’t meet therequirements. On account to human bone consisting of compact cortical bone and loosetrabecular bone is natural heterogeneous material, and the two kinds of bone have obviousboundary. Previous skeleton model and biomechanical analysis generally ignore thisdifference. But complete cortical bone and fractured trabecular bone are often seen in themedicine, so the mesh generation of monolayer isosurface heterogeneous bone has greatsignificance.To solve the mesh generation problem of two-phase bone, this paper presents a newmethod to generate a monolayer isosurface to adaptively tetrahedralize coplane heterogeneousobjects from arbitrary amount of volumetric data. One thing to notice is that the tetrahedragenerated from small tetrahedra are optimal because their smallest dihedral angle is biggestfrom the set of all possible generated tetrahedra.First, by using the method of without re-initialization level set for two-phase bone, the CTvalue of the interface we get is330Hu. Based on the CT value, we also obtain the volumerendering images of two-phase bone with implemented Ray-casting algorithms on CUDA.Second, MT(Marching Tetrahedron) algorithm is used to generate a monolayer isosurfacein the predefined background grid, then we tetrahedralize all meshes based on the isosurface.And the division of small tetrahedron is the optimal division with its biggest dihedral anglefrom the set of all possible generated tetrahedra.Final, modified mass-spring method is used to improve the quality of the generatedtetrahedra which are not boundary. Experiments are showed that the isosurface of two-phaseobject can be separated by coplane tetrahedra, and it produces more high quality tetrahedra and less boundary tetrahedra.With the tools’s help of VTK and TetGen, tetrahedron generation system from two-phaseobject has been finished. The input and output file and the function of all modules areintroduced. This paper also illustrates the overview flow chart and mesh optimization flowchart. There are the experimental mesh details of box, sphere and head data generated by thetetrahedron generation system from two-phase object.
Keywords/Search Tags:CT data, two-phase bone, adaptive, biggest dihedral angle, optimaltetrahedron
PDF Full Text Request
Related items