Related Experiment Video
Updated: Jul 16, 2026
07:51
A Novel Approach to Monitoring Graft Neovascularization in the Human Gingiva
Published on: January 12, 2019
Digital recognition of color alteration in gingiva using a convolutional neural network
Ali Raad Abdulazeez1, Hashim Mueen Hussein2, Mahmood Ali Jumaah3
1Department of Periodontics, College of Dentistry, University of Al-Mashreq, Baghdad, Iraq.
Dental and Medical Problems
|July 15, 2026
Summary
A new convolutional neural network (CNN) algorithm accurately detects gingival color changes, offering an objective digital method to assess early signs of gingivitis and improve oral hygiene awareness.
Area of Science:
- Artificial Intelligence in Dentistry
- Digital Health Technologies
- Periodontal Diagnostics
Background:
- Gingivitis is prevalent due to poor oral hygiene.
- Traditional dental indices assess periodontal status.
- Digital tools are emerging for oral health assessment.
Purpose of the Study:
- Develop an objective digital method using CNNs.
- Differentiate gingival index scores 0 and 1.
- Improve accuracy in detecting early gingival color changes.
Main Methods:
- Trained 10 CNN models on 6,660 gingival images.
- Classified images as normal (0) or abnormal (1) gingival color.
- Evaluated model performance using accuracy and Cohen's kappa.
Main Results:
- Model 4 achieved 99.9% accuracy with excellent inter-rater reliability (κ = 1.00).
- Models 8 and 5 also showed high accuracy (99.7% and 99.3%).
- All models demonstrated statistically significant agreement with expert diagnoses.
Conclusions:
- CNNs offer an objective method for detecting gingival color alterations.
- This technology can enhance patient awareness and motivation for oral hygiene.
- Potential to improve periodontal health outcomes and encourage dental visits.