This research article investigates using a deep learning model, specifically YOLOv8, to automatically detect and stage periodontitis from panoramic radiographs. The study utilized a large dataset of panoramic radiographs, which were annotated by trained dental radiologists to create a ground truth for the model's training and testing. The YOLOv8 model showed promising results in accurately identifying key anatomical structures and classifying the severity of radiographic bone loss, although some limitations, particularly with complex tooth morphologies, were noted. The authors compared YOLOv8's performance to other deep learning architectures, highlighting its superior accuracy in this context. Finally, the study concludes that this model could assist dentists in the diagnosis of periodontitis, providing a valuable clinical aid.