AI-based prediction and classification of root caries using radiographic images.

Abstract

Background: Root surface caries, commonly known as root decay, is a common dental disorder that affects tooth roots. Like enamel-based tooth decay, root caries attack exposed root surfaces caused by gum recession or periodontal disease. Older persons with gum recession, tooth loss, or poor oral hygiene may be more likely to develop this disorder. Dental root caries must be diagnosed early to improve treatment and prevention. This research will examine radiographic image-based AI-based root caries prediction algorithms.

Methods: Saveetha Dental College supplied 200 root surface radiographs. An expert dentist and dental radiologist confirmed one hundred teeth with root caries and 100 without. Edited and segmented radiographic images. Orange, a machine learning squeeze net embedding model with Naive Bayes, Logistic Regression, and neural networks, was used to assess prediction accuracy. Training and test data were split 80/20. Cross-validation, confusion matrix, and ROC analysis assessed model performance. This study examined precision and recall.

Results: Naïve bayes and logistic regression have 96% and 100% accuracy, but class accuracy is -94% and 100% in image classification of root caries was seen.

Conclusions: AI-based root caries prediction utilizing radiographic images would improve dental care by diagnosing and treating early, accurately, and personalized. With appropriate deployment, research, and ethics, AI integration in dentistry could benefit practitioners and patients. Dental professionals and AI experts must work together to maximize this new technology.AI integration in dentistry can significantly improve root caries diagnosis and treatment by predicting root caries using radiographic images. This early detection reduces treatment need and time. Collaboration between dental professionals and AI experts is crucial for maximizing benefits.