A novel method for the 3-dimensional simulation of orthognathic surgery by using a multimodal image-fusion technique

Introduction: The aim of this study was to establish a novel method for simulating orthognathic surgery in 3-dimensional (3D) space. Methods: This system mainly consists of 6 procedures: (1) reconstruction of a virtual skull model (VS) from presurgical computed tomography scans; (2) reconstruction of virtual dentition models from 3D surface scanning of dental casts occluded at presurgical and postsurgical intercuspal positions (VD1 and VD2, respectively); (3) reconstruction of a preliminary fusion model of VS and VD1 by an initial intermodality registration; (4) reconstruction of another preliminary fusion model of VS, VD1, and VD2 by a second intramodality registration; (5) repositioning of bony segments by a third intramodality registration and reconstruction of final fusion models at presurgery and postsurgery; and (6) 3D analysis of the movement of bony segments. To test this system, 2 patients with severe skeletal deformities, who had undergone presurgical orthodontic treatment, were used as models. Registration accuracy was determined by the root mean squared distance between the corresponding fiducial markers in a set of 2 images. Results and Conclusions: The sum of the root mean squared error of the 3 registration processes was less than 0.4 mm in both patients. This simulation system could be used to precisely realize the presurgical and postsurgical occlusal relationships and craniofacial morphology of a patient with severe skeletal deformities, and to quantitatively describe the movement of a given anatomical point of bony segments. It is assumed that there could be significant benefits in sharing visual and quantitative 3D information from this simulation system among orthodontists and surgeons.