Related Experiment Video
Updated: May 28, 2026

05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
A Dual-Branch Deep Learning Framework with Explainability for Dental Caries Classification Using Intra-Oral
Lijuan Ren1,2, Jinjing Chen1
1School of Software Engineering, Chengdu University of Information Technology, Chengdu 610225, China.
Journal of Imaging
|May 26, 2026
Summary
Related Concept Videos
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Sort by
Same author
Homocystinuria presenting with cerebral venous thrombosis: a case report highlighting progressive thrombosis.
Frontiers in cardiovascular medicine·2026
This study introduces a new framework for detecting dental caries, improving accuracy in both photographs and radiographs using advanced AI techniques. The developed methods show strong performance and generalize to other medical imaging tasks.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Computer Vision
Background:
- Dental caries detection is challenged by image artifacts like poor illumination and low contrast.
- Existing methods struggle with modality-specific imaging issues in dental diagnostics.
Purpose of the Study:
- To develop a comprehensive framework for accurate dental caries detection.
- To address imaging challenges in intra-oral photographs and radiographs.
- To create an AI model with explainability for clinical use.
Main Methods:
- HybridAugment+: An entropy-guided adaptive data augmentation strategy.
- DBAttNet: A dual-branch attention network with specialized attention mechanisms (IRAA for photos, CFA for radiographs).
- CAM-based explainability: Systematic evaluation and selection of the optimal method (XGrad-CAM).
Main Results:
- HybridAugment+ improved performance by up to 8.72% (photos) and 7.67% (radiographs).
- DBAttNet achieved high F1-scores: 97.90% (photos) and 95.72% (radiographs), outperforming other networks.
- XGrad-CAM identified as the best explainability method with optimal thresholds.
Conclusions:
- The proposed framework significantly enhances dental caries detection accuracy.
- The AI model demonstrates robust performance and generalizability across different medical imaging modalities.
- The study provides a foundation for AI-driven diagnostic tools in dentistry and beyond.