YOLOv10-based deep learning model for dental plaque object detection in intraoral photographs of orthodontic patients.

BACKGROUND: Orthodontic patients undergoing fixed appliance treatment often find it challenging to observe plaque due to the presence of intraoral brackets, archwires, and other devices. Consequently, inadequate brushing leads to plaque retention, triggering issues such as enamel demineralization, dental caries, and periodontal disease. Therefore, this study systematically evaluates the performance and feasibility of the YOLOv10 model in identifying dental plaque from intraoral photographs that were taken by fixed orthodontic patients using their smartphones. METHOD: This study collected intraoral plaque staining images from 23 patients undergoing fixed orthodontic treatment at the Orthodontics Department of Jilin University Stomatological Hospital between March 2024 and June 2025. We categorized these images into three sets: training, validation, and test sets. Using a GPU-equipped computer, we installed Python, PyTorch, Torchvision, and other necessary software packages to configure the environment for YOLOv10. We trained the model on the training set, validated it on the validation set, and tested it on the test set to determine the optimal learning weights, enabling the algorithm to identify plaque in intraoral photographs of orthodontic patients. RESULTS: Experimental results show that the model achieves a plaque identification precision of 89.95% for patients undergoing fixed orthodontic treatment, with a recall rate of 86.68%, an F1 score of 86.98%, an mAP50 of 93.82%, and an mAP50-95 of 64.67%. The model’s detection success rate significantly outperforms the visual identification of plaque images by both patients and orthodontists. CONCLUSION: The aforementioned results demonstrate that YOLOv10 achieves high precision and real-time dental plaque detection without the need for staining agents, exhibiting promising prospects for clinical translation. This study enables fixed orthodontic patients to obtain real-time plaque detection results simply by taking intraoral photos with a smartphone. In the future, integrating the model into software will significantly enhance the feasibility and effectiveness of daily oral hygiene management for patients undergoing fixed orthodontic treatment.