Reproducibility of a three-dimensional skeletal-based craniofacial orientation method for virtual surgical planning

Abstract

Inadequate craniofacial orientation of computed tomography (CT) scans can have significant implications in all three planes of space. The purpose of this study was to present the reproducibility of a 3-dimensional skeletal-based method of craniofacial orientation for virtual surgical planning. The protocol was defined by landmarks commonly used for cephalometry, and required identification of basion, nasion, right porion, and right orbitale, and navigation in all CT views (coronal, sagittal, and axial) for correction of yaw, roll, and pitch. Reproducibility of the method was assessed using eight CT scans that were randomly selected and anonymised. The observer group consisted of six oral and maxillofacial surgeons with varying levels of experience (resident or faculty) who performed craniofacial orientation according to the proposed method. Results were expected to be below 2° of variation, when overall accuracy as well as the influence of the academic level of the observers and symmetry of the evaluated anatomy, were considered as independent variables. Overall accuracy for all cases and for yaw, roll, and pitch were always below 2° of variation, without influence of level of experience and symmetry. Interobserver assessment was categorised as excellent in all instances, and intraobserver evaluation demonstrated consistency in the orientation of all axes. The proposed craniofacial orientation protocol presented in this study is easy to learn, applicable to computer-aided surgical planning, and can be performed by the non-technical clinician, resulting in excellent reproducibility and consistency.