Enhanced Cervical Vertebral Maturation Assessment Using AI-Based Imaging Analysis System

This study addresses dental malocclusion, which affects a significant portion of adolescents and adults and can impair both function and aesthetics. Recognizing the timely intervention importance, particularly in adolescents, this research emphasizes the role of es-timating skeletal growth in orthodontic treatment. While traditional methods rely on hand-wrist radiographs and cervical vertebral mat-uration (CVM) staging, our system uses lateral cephalometric images to classify CVM stages accurately. The system integrates a multistage classification process that combines contour, first-order, and second-order features for improved interpretability and performance. By utilizing a dataset of 509 cephalometric images, our method incorporates explainability in its decision-making process and features a user-friendly interface for clinical staff. The results indicate that the system achieved exceptional accuracy rates of up to 99.23% across multiple classifiers, demonstrating its potential to enhance CVM stage prediction and facilitate a more efficient workflow in orthodontic analysis.