Algorithms of 3D segmentation and reconstruction based on teeth CBCT images

Abstract

3D CBCT imaging plays an important role in diagnosis and treatment planning in orthodontics. The 3D tooth segmentation and reconstruction are a vital precondition. In the article, we propose two algorithms: (1). a novel image segmentation algorithm with two thresholds to separate teeth from soft tissue; (2). a local region-growing algorithm for 3D tooth extraction is created by using the local maximum gray-scale value taken as the seed point. The resulting object is defined as a single connected component containing both the seed point and voxels with gray-values lying between the local seed point and the threshold. The 3D tooth is reconstructed from the stack of the segmented image slices. Then we use VC++, OpenGL, VTK and ITK to implement the algorithms, automatic rearrangement of teeth, and 3D virtual simulation of treatment planning. It has been shown that the algorithms are robust and effective and the 3D teeth visualization software can fit into dentist's diagnosis and treatment. © 2011 Springer-Verlag Berlin Heidelberg.