Digital analysis of dental x-rays

Abstract

Digital image analysis is now considered an importantcomplementary tool for medical diagnosis in a wide varietyof instances. Several pathologies can be detected and classified with computer aided support. Dentistry uses the dental x-rays as an aided in their diagnoses. In this paper we describe several digital image analysis techniques focused on the identification of teeth tontours. Contour information is then processed in order to provide quantitative parameterization (ifthe degree of periodontitis). In our work we implemented three techniques of analysis. The first method uses the gradient information, the second method uses probabilistic models of the intensity distribution, and finally, the third method uses the Deformable models (GVF Snakes). Special attention is given to the initialization modes used in each technique. Finally we show the experimental results of each method regarding the extractiun of the crown contours and the detection of the gum level.