Detection and Classification of Dental Caries in Cone-Beam Computed Tomography

Detection and Classification of Dental Caries in Cone-Beam Computed Tomography

somayyeh imanali zadeh1 hirad tajvar2

1) Iran University Faculty of artificial intelligence of south Tehran Email:
2) Iran university Faculty of engineering and technology of lahidjan Email:

IOI Code : XKBA-AAAABG
Publication : 3rd International Conference on Recent Advances in Engineering, Innovation and Technology - Belgium(eitconf.com/3rd)
Abstract :
    the present study was an attempt evaluate the accuracy of deep learning algorithms in detecting dental caries and classifying the extent and location of dental caries in cone-beam computed tomography (CBCT) images. To the best of our knowledge, the present study represents the first investigation into using deep learning for dental caries detection in CBCT images. The dataset included CBCT scans of 382 molar teeth affected by caries and 403 unaffected molar teeth. The dataset was divided into a development set for training and validation and a test set. For each case, three images—axial, sagittal, and coronal—were obtained. The test dataset was inserted into a multi-input convolutional neural network (CNN). The network predicted the presence or absence of dental caries and classified the lesions based on their depth and type for the provided samples. Metrics such as accuracy, sensitivity, specificity, and F1 score were calculated for the detection and classification of dental caries. For carious molars, the diagnostic accuracy, sensitivity, specificity, and F1 score were 93.2. %, 96.3%, 92.1%, and 95.3%, respectively. For noncarious molars, these metrics were 94.6%, 95.8%, 94.3%, and 94.8%, respectively. The convolutional neural network demonstrated high sensitivity, specificity, and accuracy in classifying the extent and location of caries. These findings highlight the potential of deep learning models to detect and classify dental caries with high precision, sensitivity, and specificity. The successful integration of deep learning in this field could undoubtedly enhance diagnostic accuracy and treatment planning, benefiting both dental practitioners and patients.

Keywords : dental caries computed tomography artificial intelligence deep learning