Analysis of 3-D maxillofacial image data

Abstract

With rapid advances in medical imaging technology, the use of 3-D maxillofacial image data for orthodontic treatment and maxillofacial surgery is becoming more widespread. This thesis presents novel methods to analyze maxillofacial image data for the purposes of orthodontic treatment planning and maxillofacial surgery. The three-dimensional data is obtained as surface data in the form of digitized dental plaster models and as volume data in the form of computed tomography (CT) images. We first devise a novel technique for the detection of edges and tubular structures in 3-D image data. Algorithms are developed to transform the surface data to range images and to detect image features by analyzing gradient orientations. Dental features are extracted from the range images, which then leads to the automatic determination of the dental arch and tooth segmentation. The technique is also applied to CT data for extracting the nerve canal in the mandible (lower jawbone).