Enhanced Dental Disease Detection with YOLOv8: A Deep Learning Approach

The study's objective is to introduce an advanced caries detection system uses deep learning methods to evaluate panoramic dental X-ray images. The YOLOv8 framework is customized to improve detection accuracy through structural modifications in the backbone, neck, and head. Preprocessing techniques are employed to enhance detection efficiency, and the dataset, meticulously annotated by dental professionals, undergoes augmentation with random flipping and rotation to improve generalization. Precision, recall, and mean Average Precision (mAP) are used to evaluate the effectiveness of the system. Results demonstrate a notable enhancement in caries detection accuracy compared to conventional approaches. Furthermore, a confusion matrix analysis is conducted to evaluate model predictions and identify misclassification patterns.