Possibilities of artificial intelligence use in orthodontic diagnosis and treatment planning: Image recognition and three-dimensional VTO

Abstract

Orthodontic diagnosis is a comprehensive procedure that integrates various information obtained from the facial and occlusal structure as well as patient's individual needs. Hence it is not easy to imagine if the artificial intelligence(AI) would eventually replace the conventional diagnostic process. Nonetheless, recent advances in the machine learning and artificial intelligence have been applied to the cephalometric tracing and model analysis via automated image recognition, exhibiting relatively high reliability. Based on the cumulated experiences and research outcomes, orthodontic diagnostics have taken a small step towards an automated process. Considering that the orthodontic diagnosis starts from the recognition of the space discrepancy between the initial status and idealized occlusion, semi-automated three-dimensional visualized treatment objectives (VTO) may be established. This article covers the brief overview of the current status in machine learning especially focusing on the image recognition. Recent advances in the fabrication of three-dimensional VTO using surface landmarks and automated setup process is demonstrated. In the near future, a more clinically relevant VTO can be utilized using the imaginary center of resistance, to provide useful clues to the orthodontists in many of the borderline cases between extraction and non-extraction, and between surgery and non-surgery.