American journal of dentistry · 2025
Evaluation of deep learning systems in detection of dental caries on panoramic radiography.
СтатьяНазвание и аннотация: Перевод пока не подготовлен
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Biltekin Hatice, Geduk Gediz, Altan Aytaç, Karasu Seçkin
Аннотация
PURPOSE: To evaluate the effectiveness of the deep convolutional neural network model for the detection of dental caries on panoramic radiographs.
METHODS: A total of 2660 images of healthy and decayed labeled teeth were obtained from 101 panoramic radiographs. A total of 5,000 data sets were created by obtaining 2,340 synthetic data from real data. The total dataset is randomly divided as 80% training data and 20% test data. A deep learning model was created using the ResNet50 deep convolutional neural network architecture and model performance was measured after the model training. All data was evaluated and diagnostic accuracy, sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), ROC (receiver operator characteristics) curve and AUC (area under the curve) were calculated for the detection and diagnostic performance of the deep learning method with ResNet50.
RESULTS: The deep learning model classified 500 healthy and 500 decayed tooth data at a rate of 82%. The deep learning model's PPV value was 75.8%, NPV value was 92%, sensitivity 94% and specificity 70%. The AUC value was found to be 82%.
CLINICAL SIGNIFICANCE: The deep learning model used for the detection of caries in panoramic radiography is promising for use as an auxiliary tool for dentists in clinical practice.
METHODS: A total of 2660 images of healthy and decayed labeled teeth were obtained from 101 panoramic radiographs. A total of 5,000 data sets were created by obtaining 2,340 synthetic data from real data. The total dataset is randomly divided as 80% training data and 20% test data. A deep learning model was created using the ResNet50 deep convolutional neural network architecture and model performance was measured after the model training. All data was evaluated and diagnostic accuracy, sensitivity, specificity, PPV (positive predictive value), NPV (negative predictive value), ROC (receiver operator characteristics) curve and AUC (area under the curve) were calculated for the detection and diagnostic performance of the deep learning method with ResNet50.
RESULTS: The deep learning model classified 500 healthy and 500 decayed tooth data at a rate of 82%. The deep learning model's PPV value was 75.8%, NPV value was 92%, sensitivity 94% and specificity 70%. The AUC value was found to be 82%.
CLINICAL SIGNIFICANCE: The deep learning model used for the detection of caries in panoramic radiography is promising for use as an auxiliary tool for dentists in clinical practice.