Research On Algorithms In Computer Aided Orthodontic

As computer performance improving, and three-dimensional scanning, graphics,image processing, virtual reality and other technological advancing, modern medicineand computer technology is being cross-integration, and mutual penetration, modernmedicine is facing a digital revolution．Computer Aided Orthodontic is research hotspot in orthodontics. Use advancedscanning devices to obtain patients’ digital dental models, and outputs digital model oftreatment appliance by the procedures of dental surface reconstruction, toothsegmentation, tooth movement and visual treatment plan. From the perspective ofclinical practice of teeth adjustment, this paper made a profoundly research on surfacereconstruction, tooth segmentation, teeth collision detection in virtual treatment plan,The main contributions are listed as follows:（1）We present a surface reconstruction algorithm for Scattered point cloud basedon the Delaunay Triangulation and regional growth. The algorithm combines theadvantages of the two types of reconstruction algorithm of Delaunay triangulation andregional growth, meanwhile, produces better efficiency and reconstruction effects.The experiment shows that the algorithm solves the dental model reconstruction forintricate details of and complex topology dental point clouds.（2）We present an interactive segmentation algorithm for digital dental meshmode used based on geodesic path. Pick up points around traget tooth on dental meshmodel interactively, and then compute the closed geodeisc path, segment individualtooth form dental model using region growth. The experiment indicates that theproposed interactive segmentation algorithm can separate individual tooth fromcomplicate model, including serious over-crowded and edentulous case, while theborder of the segmented tooth is smooth and meets the clinical requirement oforthodontics.（3）We present a fast collision detection algorithm based on improved OBB’s indigital tooth movement. By introducing the area of a triangle as weights, overcomethe impact of the vertex sampling uneven for the tightness of OBB bounding box, andimprove the efficiency of collision detection. The experiments show that: the collisiondetection algorithm meet the realism of computer virtual teeth arrangement and realtime requirements.