Cleft, impacted teeth and maxilla lesion, are some of prevalent skeletal disorders in maxillofacial areas. In most of people, asymmetric parts of the body, imply that there is a defect in that area. In this paper we propose a new method based on symmetry detection, to find the asymmetric areas for locating mentioned disorders in cone beam computed tomography. Scaleinvariant feature transform, is a strong way to determine symmetry line of the images. After defining such line, using Bspline registration method, we compare the two sides of the image and segment the disorder. This method has been implemented on 30 real images of maxillofacial area and has shown desirable performances.
