Abstract
BACKGROUND AND AIMS: Dentofacial orthopedic treatment planning highly depends on the estimation of skeletal growth peak. In most cases, chronological age and biological age differ, some techniques estimate skeletal age by analyzing cervical vertebrae maturation (CVM) staging on lateral cephalograms. In this study, we aimed to review the different applications of AI in orthodontics and specifically discussed the different designs of AI models used for CVM estimation. METHODS: Comprehensive searches was conducted across databases including PubMed, Web of Science, Google scholar, Embase, and Scopus using keywords such as orthodontics, cervical vertebra maturation, skeletal age and artificial intelligence. RESULTS: Utilizing AI algorithms in assessing CVM-based skeletal age enhanced the accuracy of diagnosis, reduced analysis time and minimized observer variability. Deep learning techniques, especially convolutional neural networks (CNNs), have demonstrated promising results in identifying CVM stages from lateral cephalometric radiographs. CONCLUSION: AI algorithms can assist orthodontists in identifying CVM stages on lateral cephalograms. While the accuracy of AI algorithm depends on factors such as data set size, labeling methods, and model design, expert supervision and a solid understanding of CVM principles remain essential. AI should be considered as a supportive tool, not a replacement for clinical judgment.