Automatic localization of supraorbital and infraorbital foramina region on CBCT images

Abstract

CBCT (Cone Beam Computed Tomography) has fast become a key to produce 3D images in orthodontic and maxilla-facial surgery. Our purpose in this study is to develop a new approach for automatic localization of the region of supraorbital and infraorbital foramina using CBCT images. 19 CBCT images of Moroccan patients were used to validate the proposed algorithm. The automatic algorithm is based on geometric and anatomic information of maxillo-facial structures. The local priori information of supraorbital and infraorbital foramina was used to design the automatic algorithm. The detection rate for supraorbital foramina was 90% and 85% for right and left sides respectively, while the detection rate for infraorbital foramina was 79% and 74% for right and left sides respectively. This study has shown the relevance of local criteria such as geometric and anatomic knowledge in the automation of localization of supraorbital and infraorbital region. This work has gone some way towards the automation of morphometric and anthropometric measurements using CBCT images.