Categories
Change Password!
Reset Password!
Artificial intelligence-assisted caries detection tool shows strong potential for improving caries diagnosis in clinical settings. However, its sensitivity varies by tooth location and type of carious lesion.
In a recent study, researchers explored the performance of a deep learning-based artificial intelligence (AI) diagnostic model for detecting dental caries (tooth decay) through intraoral images in real-term clinical settings. The study demonstrated promising results in terms of accuracy, specificity, and predictive value, while also identifying areas for potential improvement in caries detection. Overall, 191 consecutive patients visiting an endodontics clinic were included, resulting in the examination of 4,361 teeth using an intraoral camera.
The AI model used in the study incorporated MobileNet-v3 and U-net architectures, designed to scrutinize the images and detect caries. The performance was estimated based on standard metrics: overall accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). The clinical diagnosis made by endodontic specialists was employed as the reference standard for comparative assessment. In this prospective clinical study, the AI-assisted caries detection model demonstrated strong performance in several key areas:
Overall Accuracy: The AI model attained an overall diagnostic accuracy of 93.40%. This indicates that the AI tool correctly spotted caries in the majority of cases, showing its potential to assist clinicians in identifying dental decay.
One of the key insights was the variation in diagnostic accuracy based on:
With high accuracy, specificity, and NPV, the AI tool can assist clinicians in identifying healthy teeth and successfully ruling out caries, thus boosting diagnostic confidence. However, the sensitivity for certain tooth positions and caries types needs improvement. Advanced AI models and multimodal data integration may boost AI-supported caries diagnosis in dental care settings.
BMC Oral Health
Diagnostic accuracy of artificial intelligence-assisted caries detection: a clinical evaluation
Jing-Wen Zhang et al.
Comments (0)