This research paper details the development and validation of an artificial intelligence (AI) model for automated diagnosis and classification of temporomandibular joint degenerative joint disease (TMJ DJD) using cone beam computed tomography (CBCT) images. The AI model, based on the YOLOv10 algorithm, identifies TMJ DJD and four key radiographic signs (erosion, osteophytes, sclerosis, and subchondral cysts) with high accuracy. The study involved a large dataset of CBCT images, demonstrating the model's superior performance compared to previous AI approaches and manual diagnosis by radiologists. The findings suggest the AI model could significantly improve the speed and accuracy of TMJ DJD diagnosis, aiding in early intervention. However, limitations remain regarding data imbalance and generalizability, necessitating further development and testing.