Automatic Localization of Condylar Process Centre Based on Generalized Hough Transform from CBCT images

Abstract

Cone Beam Computed Tomography (CBCT) is a technique for capturing maxillofacial images that provide tissue views with low radiation doses. In this paper, we propose an automatic method to locate the centre of the mandibular condylar process in CBCT images. The bone edges of the condylar process are extracted to detect the mandibular ramus in each slice based on the generalized Hough Transform. We propose an upward region-growing method to segment the condylar and coracoid processes. The centre of the condylar process is then determined as the centre of the condylar region in the axial slice with the maximum condyle area. The centre located by the proposed method was compared with the centre manually located by two experts. Our results indicate that the proposed method can accurately locate the condylar process centre.