Periodontal Diagnosis and Prognostication Detection Using Medical Image Processing

Abstract

Background: The elderly population faces a growing burden of various diseases, with dental issues—especially periodontal disease—often overlooked due to their asymptomatic nature. Periodontitis, however, is linked to numerous systemic conditions, leading to serious complications and negatively impacting quality of life. Affecting over a billion people globally, periodontal diseases pose a significant public health challenge due to their potential for severe oral complications. Early and accurate diagnosis is crucial, yet current methods, which rely on clinical exams and radiographs, have limitations. This study aims to develop and validate AI-driven models to enhance diagnostic accuracy and consistency in detecting periodontal disease. Methods: We analyzed 2,000 panoramic radiographs using image processing techniques. The YOLOv8 model segmented teeth, identified the cemento-enamel junction (CEJ), and quantified alveolar bone loss to assess stages of periodontitis. Results: The teeth segmentation model achieved an accuracy of 97%, while the CEJ and alveolar bone level segmentation models reached 98%. Our AI model demonstrated a remarkable performance with 94.4% accuracy and perfect sensitivity (100%). In comparison, periodontists achieved 91.1% accuracy with a sensitivity of 90.6%. General practitioners (GPs) also benefited from AI assistance, achieving 86.7% accuracy and 85.9% sensitivity, with AI enhancing diagnostic outcomes further. Conclusions: This research underscores the transformative potential of AI in dental diagnostics, highlighting its crucial role in reducing diagnostic errors, saving time, enhancing patient care, and optimizing healthcare efficiency. The implications are profound, suggesting that AI integration in periodontal diagnostics may become standard practice, significantly improving patient outcomes and streamlining dental care processes.

-