Caries Detection on Intraoral Images Using Artificial Intelligence

Abstract

Although visual examination (VE) is the preferred method for caries detection, the analysis of intraoral digital photographs in machine-readable form can be considered equivalent to VE. While photographic images are rarely used in clinical practice for diagnostic purposes, they are the fundamental requirement for automated image analysis when using artificial intelligence (AI) methods. Considering that AI has not been used for automatic caries detection on intraoral images so far, this diagnostic study aimed to develop a deep learning approach with convolutional neural networks (CNNs) for caries detection and categorization (test method) and to compare the diagnostic performance with respect to expert standards. The study material consisted of 2,417 anonymized photographs from permanent teeth with 1,317 occlusal and 1,100 smooth surfaces. All the images were evaluated into the following categories: caries free, noncavitated caries lesion, or caries-related cavitation. Each expert diagnosis served as a reference standard for cyclic training and repeated evaluation of the AI methods. The CNN was trained using image augmentation and transfer learning. Before training, the entire image set was divided into a training and test set. Validation was conducted by selecting 25%, 50%, 75%, and 100% of the available images from the training set. The statistical analysis included calculations of the sensitivity (SE), specificity (SP), and area under the receiver operating characteristic (ROC) curve (AUC). The CNN was able to correctly detect caries in 92.5% of cases when all test images were considered (SE, 89.6; SP, 94.3; AUC, 0.964). If the threshold of caries-related cavitation was chosen, 93.3% of all tooth surfaces were correctly classified (SE, 95.7; SP, 81.5; AUC, 0.955). It can be concluded that it was possible to achieve more than 90% agreement in caries detection using the AI method with standardized, single-tooth photographs. Nevertheless, the current approach needs further improvement.

New Caries Diagnostic Tools in Intraoral Scanners: A Comparative In Vitro Study to Established Methods in Permanent and Primary Teeth

“…For a useful implementation of caries diagnostics with intraoral scanners in routine patient care, it would be desirable if the intraoral scanner software would automatically mark conspicuous areas in the intraoral scan with the help of artificial intelligence. This has already been described in the field of reporting radiographs [ 57 ], clinical photos [ 58 ], or NIRI images [ 59 ]. Finally, these results must be clarified by a prospective clinical trial, which is in preparation.…”

A critical analysis of research methods and experimental models to study pulpitis

“…mutans evolution in dental caries (Aminoshariae et al, 2021; Gao et al, 2021; Javed et al, 2020; Lin et al, 2021; Zheng et al, 2021). Artificial neural networks were trained on intraoral images or radiographs to predict the presence and the evolution of caries responsible for pulpitis and showed accurate performance (Kühnisch et al, 2021; Lee et al, 2021). Interestingly, the performance of pulpitis‐predictive approaches was enhanced by adding clinical data such as spontaneous pain or tooth location, suggesting that AI could benefit from the combination of different types of data (Zheng et al, 2021).…”

Attenuation of near-ultraviolet, visible and near-infrared light in sound and carious human enamel and dentin

“…The ability to distinguish between sound and carious tissue, especially within dentin close to the pulp, is an essential requirement for the development of future caries diagnostic systems, where artificial intelligence may be able to contribute to automated and objectified diagnostics and ultimately therapy [ 24 – 26 ].…”