Automatic Identification Of Landmarks In X-Cephalometrics Analysis And Simulation Of Craniofacial Surgery

With continuously development of the calculator technique, X-Cephalometrics Analysis and Craniofacial Surgery Simulation system often be used in clinic. Automatically locating landmarks is the key of automatically accomplishing the analysis and reducing errors by hand-measurements.We apply many algorithms due to landmarks that have different characters in the lateral cephalogram. Some template matching algorithms can recognize isolated point. Some edge detection algorithms can extract boundaries, and then landmarks can be recognized on the basis of the knowledge of the human facial structure and charters of boundaries. Firstly, we locate the initial position of boundary by affine transformation that uses four acquired points (four key landmarks). Then we adjust the boundary by template matching based on affine transformation, which can avoid invalidation of template matching due to differences and distortions of target. Finally, we adjust the boundary again which tracks along the acquired boundary. To get the contour of soft tissue, we depend on connection of several templates. Lastly pay attention to the veracity of automatic identification.By inputting of CT images, 3-D reconstruction and surface rendering, the three-dimensional model of craniofacial anatomic structures were reconstructed. Through interactive operation ,we completed 3D authentic measurements including length, angle, area, volume.We use two ways to Locate Landmarks on craniofacial bone. One is projecting the volume data to the midsagittal plane and the front plane. In the planes ,the method is based what in the lateral cephalogram. Another is 3D template matching based distance transformation, for bone tissue is easily segmented in CT images. By genetic algorithm, we get the parameters of affine transformation.