Diagnostic accuracy of artificial intelligence models in detecting osteoporosis using dental images: a systematic review and meta-analysis
Authors' conclusion
This systematic review corroborates the accuracy of AI in OP diagnosis using dental images.
Study context
- Population
- Studies using AI models for osteoporosis diagnosis from dental radiographs12 studies in meta-analysis
- Comparison
- AI models (overall) vs Conventional algorithms vs Deep convolutional neural networks
- Outcome
- Sensitivity and specificity of AI models for OP diagnosis
Methodological detailsrandom-effects model, Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) risk of bias assessment, Search October 2022, updated November 2023: PubMed, Scopus, Web of Science, Google Scholar, 640 records; 22 studies in qualitative analysis, 12 in meta-analysis
- Random-effects model
- QUADAS-2 risk of bias assessment
- Search October 2022, updated November 2023: PubMed, Scopus, Web of Science, Google Scholar
- 640 records; 22 studies in qualitative analysis, 12 in meta-analysis
Pooled result
As reported in the abstract
Pooled diagnostic accuracy for OP diagnosis
| Group | Sensitivity | Specificity |
|---|---|---|
| AI models (overall) | 0.85 (95% CI, 0.70-0.93) | 0.95 (95% CI, 0.91-0.97) |
| Conventional algorithms | 0.82 (95% CI, 0.57-0.94) | 0.96 (95% CI, 0.93-0.97) |
| Deep convolutional neural networks | 0.87 (95% CI, 0.68-0.95) | 0.92 (95% CI, 0.83-0.96) |
Sensitivity
AI models (overall)0.85 (95% CI, 0.70-0.93)
Conventional algorithms0.82 (95% CI, 0.57-0.94)
Deep convolutional neural networks0.87 (95% CI, 0.68-0.95)
Specificity
AI models (overall)0.95 (95% CI, 0.91-0.97)
Conventional algorithms0.96 (95% CI, 0.93-0.97)
Deep convolutional neural networks0.92 (95% CI, 0.83-0.96)
Original publication
Diagnostic accuracy of artificial intelligence models in detecting osteoporosis using dental images: a systematic review and meta-analysis. Osteoporos Int. Published 23 August 2024. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →