article · Algorithms
Dental panoramic radiographs are often difficult to segment automatically because of high noise, low contrast, indistinct tooth boundaries, and intensity overlaps between teeth and surrounding oral tissues. To resolve these issues, a dual-stage hybrid neural network architecture named DoubleTransU-Net combines convolutional neural networks with transformers. The system uses a two-stage approach where an initial segmentation is generated first, followed by a refinement stage that clarifies ambiguous boundaries. Multi-scale context is captured using Atrous Spatial Pyramid Pooling, while squeeze-and-excitation blocks recalibrate features across channels. The architecture was tested on two public panoramic dental X-ray datasets comprising 1,500 and 1,000 images. Across both benchmarks, it achieved high precision, recall, and Dice coefficients, outperforming several existing state-of-the-art segmentation models and demonstrating reliable boundary delineation.
Accurately identifying and separating individual teeth in dental panoramic X-rays is critical for automated digital dentistry. However, poor image contrast and blurry edges frequently hinder automated tools. By successfully combining local image processing with wider contextual understanding, this architecture improves the precision of computerised tooth mapping, offering stronger technical foundations for automated dental imaging systems.
This software model is intended for automated tooth semantic segmentation, which could serve software developers building digital dentistry, diagnostic, or clinical planning tools. The technology sits at an applied and tested stage, having been validated retrospectively on two standard public benchmark datasets containing 2,500 images total, but the abstract does not report real-world clinical workflow integration.
AI-generated from the published abstract. Always read the original work before citing.
Accurate teeth segmentation in panoramic dental radiographs remains a challenging task due to high image noise, low contrast, the similarity in intensity between teeth and surrounding tissues, and blurred tooth boundaries. To address these challenges, we propose DoubleTransU-Net, a dual-stage hybrid CNN–Transformer architecture that combines progressive segmentation refinement with global contextual feature learning. The first stage generates an initial tooth segmentation, while the second stage progressively refines ambiguous tooth regions to improve boundary delineation and segmentation accuracy. In addition, Atrous Spatial Pyramid Pooling (ASPP) modules capture multi-scale contextual information, whereas squeeze-and-excitation (SE) blocks enhance discriminative feature representations through channel-wise feature recalibration. The proposed model was evaluated on two public panoramic dental X-ray datasets, UFBA-UESC (1500 images) and Tufts (1000 images), and compared against several state-of-the-art segmentation models, including U-Net, DoubleU-Net, Attention U-Net, TransUNet, and DeepLabv3+. On the UFBA-UESC dataset, DoubleTransU-Net achieved an Accuracy of 95.54%, a Dice coefficient of 93.72%, an Intersection over Union (IoU) of 88.18%, a Precision of 93.57%, and a Recall of 94.24%. On the Tufts dataset, it achieved an Accuracy of 91.83%, a Dice coefficient of 92.93%, an IoU of 86.80%, a Precision of 92.27%, and a Recall of 93.97%. These results demonstrate that DoubleTransU-Net consistently outperforms existing state-of-the-art segmentation methods while exhibiting strong robustness and generalization across different panoramic dental datasets, highlighting its effectiveness for tooth semantic segmentation in panoramic dental X-ray images.
This page summarises published work. The authoritative version sits with the publisher.
DOI: 10.3390/a19080685
Is something wrong with this record? Report it or request removal.
Discussion
Have you built on this work, tried to replicate it, or seen it applied in practice? Share what you know. Verified researchers and MARATTO™ domain experts can open a discussion, and any member can reply. Contributions are reviewed before they appear.
No discussion yet. Open the first thread.
New to MARATTO™? Create a free account.
