Artificial intelligence in orthodontics: integration with Herbst appliance therapy for diagnosis, treatment planning and outcome evaluation

Abstract

The integration of artificial intelligence (AI) into orthodontics is transforming diagnostic and therapeutic approaches, particularly in the management of skeletal Class II malocclusion with the Herbst appliance. This article reviews the application of AI-based methods — including machine learning (ML) and convolutional neural networks (CNN) — in Herbst appliance therapy, encompassing automated cephalometric landmark detection, skeletal maturation assessment, treatment outcome prediction, and post-treatment stability evaluation. A clinical study of 90 patients aged 10–16 years treated with the Herbst appliance was conducted. AI-assisted cephalometric analysis demonstrated a mean landmark detection error of 1.21±0.34 mm versus 2.18±0.61 mm by manual method (p<0.001). AI-predicted mandibular advancement outcomes matched actual results with 91.3% accuracy. The study confirms that AI serves as a powerful complementary tool in Herbst appliance orthodontic care, enhancing precision, consistency, and treatment efficiency.