3D Geometric Statistic Model’s Study And Implementation Of Spondyle

Spondylopathy, which happens with the tendency of law age, has become thecommon and high incidence diseases in our country. In recent years, the application ofthe graphics and image techniques of computer in the medical field has been improved,and3D medical image can make up the deficiency of2D, which will reduce thepossibility of misdiagnosis and diagnosis errors. However, accurate three-dimensionalpositioning and vertebral body registration is still prominent problems.In this paper, the author studies the theory of Digital Image Processing, ComputerImage Science, Statistics and theory of spine. The research work can be organized inthe following aspects.1. The establishment of3D image database of spondyle. The CT images, collectedby the medical imaging equipment, can be renewed into3D images through MCAlgorithm after the process of image restoration, correction, enhancement and edgeextraction and simplification. Then, the3D model will be put into the database.2. The calibration and pick of the spondyle’s sampling points. Throughsummarizing the relevant scholars’ definitions of cranium, craniofacial features, theauthor finished the calibration of the spondyle’s sampling points, and saved theinformation of spondyle’s sampling points into database.3. The establishment of the spondyle’s statistic geometric model. The authorestablished the spondyle’s statistic geometric model after the alignment of spondyle’ssampling points and PCA training.4. The implementation and design of Spinal3D image reconstruction system. Thecorrectness and high efficiency of the algorithms in this thesis are proven by thesystem.This subject is supported by the National Natural Science Foundation project "3Dgeometry statistic model’s studyof the spine and spondyle"(60871097).