Surgery-First Orthognathic Approach to Correct Facial Asymmetry: Artificial Intelligence–Based Cephalometric Analysis

Background: The surgery-first orthognathic approach has been applied at our institution since 2007. However, its indications remain debated. The aim of this study was to investigate the reliability of the surgery-first approach to correct facial asymmetry compared to the traditional orthodontics-first approach using a novel artificial intelligence–based cephalometric analysis. Methods: Intervention outcomes of surgery-first ( n = 33) and orthodontics-first ( n = 26) approaches to correct facial asymmetry were examined. Patients with facial asymmetry who had undergone orthognathic surgery from January of 2006 to January of 2019 were included in the study. In the surgery-first approach, the novel preoperative simulation process on the dental model was performed to determine the final occlusion without presurgical orthodontic treatment. Changes in cephalometric landmarks were compared using the supervised deep learning process developed at our institution. Results: The surgery-first approach without presurgical orthodontic treatment corrected facial asymmetry and yielded results similar to those of the traditional orthognathic approach. The statistical analysis revealed that changes in skeletal cephalometric landmarks were similar in the two groups. Conclusions: The surgery-first orthognathic approach without presurgical orthodontic treatment treated facial asymmetry, possibly suggesting a possible paradigm shift in treatment. In addition, artificial intelligence–based cephalometric analysis was an effective tool. CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, III.