Bone 3D-reconstruction Based On CT Image

With the research development of 3D medical image application in clinic, medical image 3D-reconstruction has become more and more significant. The oriented RP 3d-reconstruction method has been studied in the paper, with the focus on the bone CT image processing method and surface reconstruction with triangular facet.On the basis of study and analysis of bone CT image feature, bone contour data have been gained through the method in which the filter, segmentation, contour extraction and vector are done orderly. And the effective filter and contour extraction algorithm is proposed.Contour classifying method has been studied. The ray-intersection method is used to determine the nesting level of contours in the same layer, and the contours match relationship between two adjacent layers is established according to contours geometry comparability and relative position fixity.After research surface reconstruction with triangular facets, contours of first layer and the last layer are triangulated with layer separation algorithm firstly. And then the least polar-angle criterion is presented for linking two convex contours, and the method is also given in which non-convex contour is transformed to a set of convex contour. With the splitting and integrating algorithm, the single-to-multi-contour reconstruction with triangular facet has been simplified to one-to-one contour. In the end, the established triangle builds the whole bone.The oriented RP bone 3D-reconstruction system is developed with VC++ and OpenGL, which can realize contour extraction form CT image and bone 3D-reconstruction.The system can output various CAD interfaces, and provide data for individually matching artificial limb design and manufacture.