AI model improves periodontitis diagnosis using dental radiographs
A groundbreaking artificial intelligence (AI) model has been developed to enhance the accuracy and efficiency of diagnosing periodontitis, according to a new study published in BMC Oral Health. The two-stage AI system uses deep learning algorithms to automatically detect, stage, and grade periodontitis from panoramic dental radiographs — offering a major leap forward in automated dental diagnostics.
AI brings speed and precision to periodontitis detection
The study, conducted by Khiem Quang Do, Truc Thanh Thai, Viet Quoc Lam, and Thuy Thu Nguyen, introduces a two-stage diagnostic process combining MobileNetV2 and YOLOv8 models.
- The first stage employs the MobileNetV2 model to identify whether periodontitis is present in a radiograph, achieving an impressive 88% accuracy.
- The second stage, powered by YOLOv8, performs detailed analysis to localize bone loss and grade its severity, with testing precision and recall rates of 0.74 and 0.7, respectively.
By automating these steps, the AI system eliminates much of the subjectivity and time associated with manual interpretation, paving the way for early detection and intervention.
Quantitative, objective diagnostic insights
The model measures bone loss relative to root length and automatically assigns a stage (I–IV) of periodontitis based on standardized metrics. This allows clinicians to obtain consistent, objective data that can guide treatment planning and reduce diagnostic variability between practitioners.
The researchers emphasized that this AI tool could serve as a valuable diagnostic aid, especially for general dental practitioners who lack specialized training in periodontology. It acts as a second opinion, minimizing missed diagnoses caused by human error or fatigue.
Clinical significance and future outlook
According to the study’s authors, integrating such AI-assisted diagnostic systems into dental practice could significantly enhance precision, consistency, and patient outcomes. Beyond detection, the model may also help in developing personalized treatment strategies by offering a clearer picture of disease progression.
However, the authors also highlighted certain limitations. Broader external validation using larger, more diverse datasets across populations is needed to strengthen the model’s accuracy and ensure clinical applicability worldwide.
A step toward the future Of digital dentistry
As AI continues to revolutionize dental imaging, this innovation underscores the potential for deep learning models to transform disease detection and diagnosis. The researchers believe future improvements will allow AI tools to integrate seamlessly with clinical workflows, making early periodontal disease detection faster, more reliable, and more accessible globally.
Stay informed with Dental News Pakistan
CLICK HERE TO JOIN our WhatsApp community for verified global oral health updates, new innovations, expert insights, and breaking dental news delivered straight to your phone.