Automated assessment of peri-implant disease severity by deep learning and image processing in periapical radiographs.
Аннотация
BACKGROUND/PURPOSE: Dental implant surgery had become a standard treatment option for oral rehabilitation. The severity of peri-implant bone loss was a critical clinical indicator for evaluating the success of dental implants. This study assessed an automated framework combining deep learning and image processing techniques for classifying the severity of peri-implant bone loss using periapical radiographs and providing diagnostic assistance.
MATERIALS AND
METHODS
A total of 780 periapical radiographs containing 1210 implants were analyzed. A YOLO-based object detection model was employed to localize peri-implant regions accurately. Subsequently, we applied our custom-developed peri-implant image processing pipeline and alveolar crest localization algorithm to categorize each implant into one of three severity levels. The clinical feasibility of the proposed framework was also evaluated.
RESULTS
On a test dataset of 120 periapical radiographs, the YOLOv8-S model achieved a detection precision of 98.1 %, with a sensitivity of 96.0 % and a specificity of 99.1 %. For the three-grade classification of peri-implant bone loss severity, the highest accuracy reached 96.61 %, with an overall classification accuracy of 95.8 %. In clinical feasibility testing, our framework demonstrated a 36-fold improvement in assessment speed and approximately a 7.5 % increase in diagnostic accuracy compared to manual evaluation by dental experts.
CONCLUSION
The proposed automated framework for assessing peri-implant bone loss severity in periapical radiographs shows strong potential to assist clinical dental practices. It is a reliable second opinion to support clinical decision-making and improve diagnostic efficiency.
Подготовка и перевод публикации
Обработка статьи · этапы, ошибки и управление
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Тип публикации указан по данным источника. Уровень доказательности автоматически не присваивается.