Related Experiment Video
Updated: Aug 14, 2026
05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
A Genetic Algorithm-Optimized ConvNeXtV2-YOLOv8 Framework for Dental Caries Detection in Panoramic Radiographs
Nebras Sobahi1, Deniz Bora Küçük2, Kazım Kılıç3
1Department of Electrical and Computer Engineering, Faculty of Engineering, King Abdulaziz University, Jeddah 21589, Saudi Arabia.
Diagnostics (Basel, Switzerland)
|August 13, 2026
Summary
This study improved dental caries detection in panoramic radiographs using an enhanced YOLOv8 model. The new framework shows promise for computer-aided diagnosis, aiding dentists in identifying cavities more accurately.
Area of Science:
- Artificial Intelligence in Dentistry
- Medical Imaging Analysis
- Computer-Aided Diagnosis
Background:
- Dental caries is a prevalent global oral disease requiring early detection.
- Diagnosing caries in panoramic radiographs is difficult due to image quality and anatomical challenges.
Purpose of the Study:
- To develop an advanced object detection framework for precise dental caries localization in panoramic radiographs.
- To enhance feature extraction and gradient propagation for improved detection accuracy.
Main Methods:
- A modified YOLOv8 architecture integrated ConvNeXtV2 blocks into C2f modules.
- A genetic algorithm optimized key training hyperparameters to reduce overfitting.
- The model was trained and evaluated on 474 expert-annotated panoramic dental radiographs.
Main Results:
- The optimized model achieved 78.4% precision and 62.6% AP50 for caries localization.
- The ConvNeXtV2-enhanced architecture demonstrated superior performance compared to the baseline YOLOv8.
- Accurate localization of carious regions was confirmed through visual and quantitative analyses.
Conclusions:
- Combining ConvNeXtV2 architectural enhancements with genetic algorithm optimization effectively localizes dental caries.
- The framework shows potential as a computer-aided diagnostic tool for clinical caries assessment, despite recall limitations.
