Related Experiment Video
Updated: Aug 5, 2026
05:49
Reliability of Artificial Intelligence-Based Cone Beam Computed Tomography Integration with Digital Dental Images
Published on: February 23, 2024
Diagnostic Performance of Artificial Intelligence Models for Periodontitis Disease Detection Using Panoramic
Khalid Almutairi1, Tariq Almanseer1, Enrique España Guerrero1
1Department of Stomatology, Section of Periodontology, Faculty of Dentistry, University of Granada, Colegio Máximo s/n, Campus Universitario de Cartuja, 18071 Granada, Spain.
Dentistry Journal
|July 27, 2026
Summary
Artificial intelligence (AI) shows promise in detecting periodontitis from panoramic radiographs, achieving high sensitivity and AUC values. However, limited external validation and varied data quality necessitate AI as a supplement to clinical exams.
Area of Science:
- Dentistry
- Medical Imaging
- Artificial Intelligence
Background:
- Periodontitis is a common inflammatory disease causing tooth loss.
- Accurate diagnosis relies on clinical and radiographic data, but radiograph interpretation varies.
- Artificial intelligence (AI) offers potential for improved radiographic assessment.
Purpose of the Study:
- To systematically review the diagnostic performance of AI models for periodontitis detection on panoramic radiographs.
- To assess the effectiveness of AI in analyzing radiographic findings for periodontitis.
Main Methods:
- Systematic search of PubMed, Scopus, and Web of Science (Jan 2015–Mar 2026).
- Inclusion of studies evaluating AI for periodontitis detection on panoramic radiographs with clinical or expert annotation as reference.
- Data extraction and quality assessment using QUADAS-2, followed by narrative synthesis due to heterogeneity.
Main Results:
- Nine studies (over 20,000 radiographs) utilized AI models like CNNs.
- Sensitivity ranged from 0.795–1.00, specificity from 0.784–0.99, and AUC from 0.843–0.967.
- Performance varied based on reference standards; limited external validation and inconsistent dataset quality were noted.
Conclusions:
- AI models show promising diagnostic performance for periodontitis on panoramic radiographs.
- Heterogeneity and limited validation hinder generalizability; AI should augment, not replace, clinical examination.
- Standardized datasets and robust external validation are crucial for clinical AI implementation.
