Detecting the presence of supernumerary teeth during the early mixed dentition stage using deep learning algorithms: A pilot study

Abstract

Background: Supernumerary teeth are a common anomaly and are frequently observed in paediatric patients. To prevent or minimize complications, early diagnosis and treatment is ideal in children with supernumerary teeth. Aim: This study aimed to apply convolutional neural network (CNN)-based deep learning to detect the presence of supernumerary teeth in children during the early mixed dentition stage. Design: Three CNN models, AlexNet, VGG16-TL, and InceptionV3-TL, were employed in this study. A total of 220 panoramic radiographs (from children aged 6 years 0 months to 9 years 6 months) including supernumerary teeth (cases, n = 120) or no anomalies (controls, n = 100) were retrospectively analyzed. The CNN performances were assessed according to accuracy, sensitivity, specificity, receiver operating characteristic (ROC) curves, and area under the ROC curves for a cross-validation test dataset. Results:The VGG16-TL model had the highest performance according to accuracy, sensitivity, specificity, and area under the ROC curve, but the other models had similar performance. Conclusion:CNN-based deep learning is a promising approach for detecting the presence of supernumerary teeth during the early mixed dentition stage.

Analysis of the feasibility of using deep learning for multiclass classification of dental anomalies on panoramic radiographs

“…We previously compared the performance of three deep CNN models (i.e., AlexNet, VGG16, and InceptionV3) on the classification of supernumerary teeth. In this previous study, the classification of supernumerary teeth by these three deep CNN models yielded similar performance metrics 19) . AlexNet is an epoch-making deep CNN model that was published in 2012 21) .…”

“…When it comes to AI diagnosis, sensitivity is a crucial metric because high sensitivity means few false negatives (i.e., if a person has a negative test, they are likely to be free of the disease). However, previous studies have shown that any algorithm's performance (sensitivity) for both object detection and classification tasks for supernumerary teeth is in the 80% to low 90% range, and is not perfect for other metrics as well 19,[28][29][30] . Therefore, we presume that more stable results might be achieved by combining classification and object detection CNNs.…”

“…AlexNet is an epoch-making deep CNN model that was published in 2012 21) . Since then, various deep CNN models have been proposed; however, AlexNet remains in use and has undergone continuous improvement 19,28,31,32) .…”

“…where TP is true positive, TN is true negative, FP is false positive, and FN is false negative (all positive and negative determinations are made using counts for case detection) 19) .…”

“…We previously reported that a convolutional neural network (CNN)-based deep learning technique demonstrated good performance in the binary classification of supernumerary teeth during the early mixed dentition stage 19) . The binary classification is the task of classifying given data into one of two classes.…”