A Data-Efficient Hierarchical Deep Learning Approach for Diagnosing Impacted Canine-Induced Root Resorption via Cone-beam Computed Tomography

Abstract

Objectives: Impacted canines can cause canine-induced root resorption (CIRR), potentially leading to root resorption and even incisor loss. The aim of this study was to usedeep learning to automatically evaluate the diagnosis of CIRR in maxillary incisors more accurately using CBCT images. Methods: A total of 50 CBCT images and 176 incisors were selected for the present study. The maxillary incisors were manually segmented from the CBCT images and labeled by two independent radiologists as either healthy or having root resorption induced by the impacted canines. We used five different strategies for training the model: A) classification using 3D ResNet50 (Baseline), B) classification of the segmented masks using the outcome of a 3D U-Net pretrained on the 3D MNIST, C) training a 3D U-Net for the segmentation task and use its outputs for classification, D) pretraining a 3D U-Net for the segmentation and transfer of the model, and E) pretraining a 3D U-Net for the segmentation and fine-tuning the model with only the model encoder. The segmentation models were evaluated using the mean intersection over union (mIoU) and Dice coefficient. The classification models were evaluated in terms of classification accuracy, precision, recall, and F1 score. Results: The segmentation model achieved a mean intersection over union (mIoU) of 0.641 and a Dice coefficient of 0.901, indicating good performance in segmenting the tooth structures from the CBCT images. For the main classification task of detecting canine-induced root resorption (CIRR), Model C (classification of the segmented masks using 3D ResNet) and Model E (pretraining on segmentation followed by fine-tuning for classification) performed the best, both achieving 82% classification accuracy and 0.62 F1-scores on the test set. These results demonstrate the effectiveness of the proposed hierarchical, data-efficient deep learning approaches in improving the accuracy of automated CIRR diagnosis from limited CBCT data compared to the 3D ResNet baseline model. Conclusion: The proposed approaches are effective at improving the accuracy of classification tasks and are helpful when the diagnosis is based on the volume and boundaries of an object. The study demonstrated that the proposed approaches improve the accuracy of medical image classification tasks.