Diagnostic performance of artificial intelligence in periapical radiography: a systematic review.
Tulin Tasdemir, Melek Tassoker, Abdullah Arpacı et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review evaluated the diagnostic accuracy of artificial intelligence (AI) models in periapical radiography, comparing their performance to human experts. Out of 544 records, 47 studies were included, indicating that AI models demonstrate high diagnostic potential, performing comparably to human experts in various tasks. However, clinical applicability is hindered by significant biases and validation issues.
Key findings
- AI models show high diagnostic potential in periapical radiography, comparable to human experts.
- 47 studies met eligibility criteria out of 544 identified records.
- Clinical applicability is limited by high risk of bias and lack of external validation.
Why it matters
Understanding the diagnostic capabilities of AI in radiography is crucial for integrating these technologies into clinical practice, potentially enhancing diagnostic accuracy and efficiency. However, awareness of their limitations is essential for safe implementation.
What to keep in mind
The review highlights a high risk of bias and lack of external validation in the included studies.