We want to make three-dimensional (3D) images available for routine use in the classroom and clinical settings.
Currently, computed tomography from in vivo as well in vitro data primarily gives two-dimensional views in the three projection planes. These two-dimensional images are currently interpreted by radiologists or by care providers using specialized software. Three-dimensional reconstruction from CT data requires processes known as segmentation and classification. Both steps allow the software to detect what voxel (data point in the 3D matrix assigned to a grey level) belongs to the object in question and what material quality is represented by it.
We work with software developers such as Volume Graphics to improve methods to generate 3D images from teeth, root canals and jaw sections. This is still a labor-intensive process requiring substantial manual editing, which we are trying to streamline. In this field we work on educational content and intelligent systems to recognize patterns in CT data sets. A future direction of work will be the development of 3D finite element models for structural analyses.