Deep Learning for Caries Detection using Optical Coherence Tomography

Abstract

AbstractEarly detection of dental caries has been one of the most predominant topics studied over the last few decades. Conventional examination through visual-tactile inspection and radiography can be inaccurate and destructive to the tooth structure. The development of optical coherence tomography (OCT) has given dentistry an alternative diagnostic technique, which has been proven by numerous studies, that it has better sensitivity, specificity, and non-invasive characteristics. The growing popularity of artificial intelligence (AI) also contributes to a more efficient and effective way of image-based detection and decision-making. However, previous studies, which have attempted to employ AI for caries assessment, did not incorporate high-quality ground truth data. Therefore, this study aims to bypass this issue and highlights the importance of high-quality data. A two-phase study was carried out to explore different methods for caries detection. Initially, the comparison of caries detection based on OCT and apical radiography by 5 experienced clinicians was conducted. Then, five convolutional neural networks (CNNs), including AlexNet, VGG-16, ResNet-152, Xception, and ResNext-101, in the scope of AI were employed to detect caries and compared with the findings of the 5 clinicians. The data was preprocessed and labeled with the ground truth corresponding to microcomputed tomography (micro-CT) with rigorous definition. The weighted Kappa statistics suggested that OCT (ϰ= .699, SD = .090) showed a higher accuracy rate than radiography (ϰ= .407, SD = .049), and CNNs (ϰ= .860, SD = .049) were rated higher than clinicians (ϰ= .679, SD = .113), both at a .05 significance level. The best result was carried out by ResNet-152, which demonstrated a high accuracy rate of 95.21% and a sensitivity of 98.85%. These findings illustrate the importance of ground truth data for AI training and the potential of deep CNN algorithms combined with OCT for diagnosing dental caries.