<![CDATA[Landmark identification on lateral cephalogram images plays a crucial role in orthodontic diagnosis and treatment planning because it forms the basis for analyzing the skeletal relationship between facial structures and teeth in the oral cavity. However, manual landmark identification is time-consuming and potentially leads to subjective errors. Therefore, Artificial Intelligence (AI)-based technology is a potential solution to improve the efficiency and consistency of analysis. This study aims to develop an automatic anatomical landmark detection system on lateral cephalogram images using the YOLOv11 algorithm. The dataset used consisted of 50 lateral cephalogram images obtained from the Radiology Installation of RSGM Andalas University and annotated according to the American Board of Orthodontics (ABO) standards. Then, augmentation was performed to obtain a total of 110 images. The model training and testing process was carried out using the YOLOv11 variant “yolo11s-pose”. The evaluation results showed an accuracy, precision, recall, and F-score of 1.0, with a mean Average Precision (mAP) of 0.995. Overall, this model shows good potential in improving the efficiency of cephalogram landmark identification, but it requires increasing the amount and variety of data for more reliable performance in clinical applications.]]>
Copyrights © 2026
