Related Experiment Video
Updated: Jun 9, 2026
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
Development of an oral cancer detection system through deep learning
Liangbo Li1,2, Cheng Pu3,4, Jingqiao Tao1,5
1Medical School of Chinese PLA, Beijing, China.
BMC Oral Health
|December 5, 2024
Summary
This study developed an AI model using a portable endoscope for oral cancer detection. The deep learning approach shows promise for identifying oral cancer from intraoral images.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Oral cancer poses a significant global health challenge.
- Early detection is crucial for improving patient outcomes.
- Current diagnostic methods can be invasive or limited in accessibility.
Purpose of the Study:
- To develop and evaluate an AI-based model for oral cancer detection.
- To utilize a portable electronic oral endoscope for capturing intraoral images.
- To assess the performance of deep learning models (U-Net, ResNet-34) in identifying oral cancer.
Main Methods:
- Collected 205 annotated intraoral images of oral cancer patients.
- Employed U-Net and ResNet-34 deep learning architectures.
- Evaluated model performance using Dice coefficient, IoU, Loss, Precision, Recall, and F1 Score.
Main Results:
- During training, Dice values reached ~0.8, Loss ~0, and IoU ~0.7.
- In testing, the model achieved a maximum Precision of 0.96.
- An F1 score of 0.58 was reached with a low confidence threshold.
Conclusions:
- Deep learning models demonstrate potential for oral cancer identification.
- AI-powered analysis of intraoral images is a promising avenue.
- Further research is warranted to refine and validate these AI tools.
