Dataset Information

Generalizability of Deep Learning Models for Caries Detection in Near-Infrared Light Transillumination Images.

ABSTRACT:

Objectives

The present study aimed to train deep convolutional neural networks (CNNs) to detect caries lesions on Near-Infrared Light Transillumination (NILT) imagery obtained either in vitro or in vivo and to assess the models' generalizability.

Methods

In vitro, 226 extracted posterior permanent human teeth were mounted in a diagnostic model in a dummy head. Then, NILT images were generated (DIAGNOcam, KaVo, Biberach), and images were segmented tooth-wise. In vivo, 1319 teeth from 56 patients were obtained and segmented similarly. Proximal caries lesions were annotated pixel-wise by three experienced dentists, reviewed by a fourth dentist, and then transformed into binary labels. We trained ResNet classification models on both in vivo and in vitro datasets and used 10-fold cross-validation for estimating the performance and generalizability of the models. We used GradCAM to increase explainability.

Results

The tooth-level prevalence of caries lesions was 41% in vitro and 49% in vivo, respectively. Models trained and tested on in vivo data performed significantly better (mean ± SD accuracy: 0.78 ± 0.04) than those trained and tested on in vitro data (accuracy: 0.64 ± 0.15; p < 0.05). When tested in vitro, the models trained in vivo showed significantly lower accuracy (0.70 ± 0.01; p < 0.01). Similarly, when tested in vivo, models trained in vitro showed significantly lower accuracy (0.61 ± 0.04; p < 0.05). In both cases, this was due to decreases in sensitivity (by -27% for models trained in vivo and -10% for models trained in vitro).

Conclusions

Using in vitro setups for generating NILT imagery and training CNNs comes with low accuracy and generalizability.

Clinical significance

Studies employing in vitro imagery for developing deep learning models should be critically appraised for their generalizability. Applicable deep learning models for assessing NILT imagery should be trained on in vivo data.

SUBMITTER: Holtkamp A

PROVIDER: S-EPMC7957685 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:ObjectivesThis study aimed to investigate the accuracy of deep learning algorithms to diagnose tooth caries and classify the extension and location of dental caries in cone beam computed tomography (CBCT) images. To the best of our knowledge, this is the first study to evaluate the application of deep learning for dental caries in CBCT images.MethodsThe CBCT image dataset comprised 382 molar teeth with caries and 403 noncarious molar cases. The dataset was divided into a development set for training and validation and test set. Three images were obtained for each case, including axial, sagittal, and coronal. The test dataset was provided to a multiple-input convolutional neural network (CNN). The network made predictions regarding the presence or absence of dental decay and classified the lesions according to their depths and types for the provided samples. Accuracy, sensitivity, specificity, and F1 score values were measured for dental caries detection and classification.ResultsThe diagnostic accuracy, sensitivity, specificity, and F1 score for caries detection in carious molar teeth were 95.3%, 92.1%, 96.3%, and 93.2%, respectively, and for noncarious molar teeth were 94.8%, 94.3%, 95.8%, and 94.6%. The CNN network showed high sensitivity, specificity, and accuracy in classifying caries extensions and locations.ConclusionsThis research demonstrates that deep learning models can accurately identify dental caries and classify their depths and types with high accuracy, sensitivity, and specificity. The successful application of deep learning in this field will undoubtedly assist dental practitioners and patients in improving diagnostic and treatment planning in dentistry.Clinical significanceThis study showed that deep learning can accurately detect and classify dental caries. Deep learning can provide dental caries detection accurately. Considering the shortage of dentists in certain areas, using CNNs can lead to broader geographic coverage in detecting dental caries.

Project description:PurposeTo quantitatively evaluate the generalizability of a deep learning segmentation tool to MRI data from scanners of different MRI manufacturers and to improve the cross-manufacturer performance by using a manufacturer-adaptation strategy.Materials and methodsThis retrospective study included 150 cine MRI datasets from three MRI manufacturers, acquired between 2017 and 2018 (n = 50 for manufacturer 1, manufacturer 2, and manufacturer 3). Three convolutional neural networks (CNNs) were trained to segment the left ventricle (LV), using datasets exclusively from images from a single manufacturer. A generative adversarial network (GAN) was trained to adapt the input image before segmentation. The LV segmentation performance, end-diastolic volume (EDV), end-systolic volume (ESV), LV mass, and LV ejection fraction (LVEF) were evaluated before and after manufacturer adaptation. Paired Wilcoxon signed rank tests were performed.ResultsThe segmentation CNNs exhibited a significant performance drop when applied to datasets from different manufacturers (Dice reduced from 89.7% ± 2.3 [standard deviation] to 68.7% ± 10.8, P < .05, from 90.6% ± 2.1 to 59.5% ± 13.3, P < .05, from 89.2% ± 2.3 to 64.1% ± 12.0, P < .05, for manufacturer 1, 2, and 3, respectively). After manufacturer adaptation, the segmentation performance was significantly improved (from 68.7% ± 10.8 to 84.3% ± 6.2, P < .05, from 72.4% ± 10.2 to 85.7% ± 6.5, P < .05, for manufacturer 2 and 3, respectively). Quantitative LV function parameters were also significantly improved. For LVEF, the manufacturer adaptation increased the Pearson correlation from 0.005 to 0.89 for manufacturer 2 and from 0.77 to 0.94 for manufacturer 3.ConclusionA segmentation CNN well trained on datasets from one MRI manufacturer may not generalize well to datasets from other manufacturers. The proposed manufacturer adaptation can largely improve the generalizability of a deep learning segmentation tool without additional annotation.Supplemental material is available for this article.© RSNA, 2020.