Investigating the Application of Artificial Intelligence in the Diagnosis of Dental Caries from Radiographic Images

In recent decades, the dental industry has witnessed tremendous developments; from the introduction of advanced units to the emergence of digital imaging. But today, we are on the verge of the greatest technological revolution in the history of this science: the introduction of dental artificial intelligence. Artificial intelligence is no longer a science-fiction concept, but a tangible tool that has come to the aid of dentists to minimize human errors and increase the accuracy of caries diagnosis to an unprecedented level. Tooth decay is one of the most common oral diseases worldwide, and its early detection plays an important role in preventing the progression of the lesion, maintaining healthy tooth structure, and reducing the need for invasive treatments. Dental radiographs, especially bitewing radiographs, are among the most important diagnostic tools for identifying caries that may not be easily visible during a direct oral examination. With the development of artificial intelligence, especially machine learning and deep learning, it has become possible to automatically analyze radiographic images and identify patterns associated with dental caries. In recent years, deep neural networks and architectures such as CNN, ResNet, and the YOLO family have been used to detect, classify, and segment carious lesions. Systematic reviews have shown that the reported performance of these systems is promising, although large differences in datasets, labeling methods, radiograph type, and risk of bias between studies make it difficult to directly compare models. A meta-analysis of artificial intelligence systems for caries detection in bitewing radiographs reported a pooled sensitivity of about ۸۷% and a pooled specificity of about ۸۹%. More recent studies also show that this technology has great potential as an adjunct diagnostic tool rather than a replacement for the dentist.

کلیدواژه ها:

Artificial intelligence ، deep learning ، dental caries ، dental radiography ، bitewing ، convolutional neural network ، automatic diagnosis ، digital dentistry