Teeth segmentation and root canal identification method for Cone-beam CT images

OData support
Supervisor:
Dr. Benyó Balázs István
Department of Control Engineering and Information Technology

Diagnostic imaging plays a very important role in medical practice, especially devices using 3D imaging technique. Cone-beam CT is one of the most recent technology in conservative dentistry. In my thesis I will present an approach to analyze these images. I examined different image processing methods for the segmentation of anatomical structures - primarily teeth and their roots - in the maxillofacial region. Our aim is to support diagnostic decisions by reconstructing and visualizing the roots of the teeth along with their root canals. Three different methods were tested. Each method used a version of a popular segmentation algorithm: region growing segmentation, fuzzy c-means clustering and watershed transformation. The most successful method was the one that used watershed transformation. This process starts from a given point in a tooth and executes its automated segmentation. We need to develop a post-process algorithm because the surface of the segmented tooth was inaccurate. The results are promising: it can be a good basis of a new decision support system.

Downloads

Please sign in to download the files of this thesis.