3D Visual Modeling Based On CT Images Of Teeth

The application of3D visualization of medical imaging technology in modern medicinehas made the3D information of the body’s internal lesions more clearly shown in front of thedoctor. It’s facilitate for the doctor to make rapid and effective medical programs of thediagnosis to meet people’s increasingly demand for medical services.3D visualizationtechnology was applied in the dental diagnosis to develop a system which facilitated thedoctor to diagnose the disease in this paper. It is an efficient, accurate and low-cost solutioncompared to traditional dental treatment and greatly meet the market demand.The following issues to be solved in the system design process are as followed. Firstly,how to obtain the point cloud data information (STL format) from the teeth DICOM image toachieve the preliminary reconstruction of the teeth for the follow-up3D CAD reversereconstruction. Secondly, how to reconstruct a realistic teeth surface model based on teethpoint cloud.Firstly, tooth preliminary reconstruction. Because of the unclear dental image boundaryand gentle gray gradient, the teeth DICOM image should be pre-processed Firstly, this papermade a comparison of median filter and Gaussian filter and found median filtering had goodeffect and fast efficiency, so used median filter image filtering process to remove noise. Thenthe processed images were processed by binarized image segmentation method to obtain theinterested region and used the marching cubes algorithm to achieve surface rendering,3Dvisualization. But the results did not really realize the digitizing and can not provide data onthe latter part of the reconstruction. To solve this problem, the results of the surface renderingof triangular facets was streamlined and smooth processed in this paper and then saved it asSTL format file to build point cloud data source for later reconstruction.Secondly, tooth curved surface model reconstruction. Because of a lot of noise anduseless points in obtained point cloud data, the point cloud data must be processed. Firstlydirect observation method was applied to delete the noise and useless points, followed byreducing data and reserving characteristics. Must reconstruct the part of point cloud pointsand raised the point cloud segmentation method based on curvature and morphologicalcharacteristics. Due to the lack of some characteristic data in various parts, applied methodbased on B-spline curve to fill the defect point cloud. Then reconstructed curved surface andevaluated reconstructed curved surface quality, thus obtained a satisfactory curved surface.After solving the above two problems, according to market demand, developed and designed3D visual modeling based on CT images of teeth and reconstruction system. Thesystem was based on Visual Studio2008and Visualization Tool Kit (VTK) and appliedobject-oriented programming techniques to achieve the development of the system which hadthe characteristics of scalability, portability and modular, thus facilitated the follow-up furtherstudy.