Related Experiment Video
Updated: May 10, 2025
05:49
Author Spotlight: Advancing CBCT and Digital Dental Image Integration with AI-Assisted Digitization
Published on: February 23, 2024
743
Decoding wisdom: Evaluating ChatGPT's accuracy and reproducibility in analyzing orthopantomographic images for third
Ana Suárez1, Stefania Arena1, Alberto Herranz Calzada1,2
1Department of Pre-Clinic Dentistry II, Faculty of Biomedical and Health Sciences, Universidad Europea de Madrid, Calle Tajo s/n, Villaviciosa de Odón, Madrid 28670, Spain.
Computational and Structural Biotechnology Journal
|April 24, 2025
Summary
ChatGPT-4o shows limited accuracy in interpreting dental radiographs like orthopantomograms (OPGs). While consistent, its current limitations require professional oversight for safe clinical integration in AI-assisted dentistry.
Area of Science:
- Artificial Intelligence in Healthcare
- Dental Radiology
- Clinical Decision Support Systems
Background:
- Artificial Intelligence (AI) integration offers potential for clinical decision support in radiology.
- Evaluating AI tools like ChatGPT-4o for specific diagnostic tasks is crucial for safe implementation.
Purpose of the Study:
- To assess the accuracy and reproducibility of ChatGPT-4o in interpreting orthopantomograms (OPGs).
- To evaluate ChatGPT-4o's performance in simulating patient requests for lower third molar extraction assessments.
Main Methods:
- Thirty OPGs were analyzed using standardized prompts submitted to ChatGPT-4o, generating 900 responses.
- Oral surgery experts evaluated responses using a three-point Likert scale; disagreements were resolved by a third expert.
- Accuracy and repeatability metrics, including Gwet's coefficient, were calculated.
Main Results:
- ChatGPT-4o achieved an overall accuracy rate of 38.44% in OPG interpretation.
- Response agreement among repetitions was 82.7%, but Gwet's coefficient indicated moderate repeatability (60.4%).
- The model frequently provided incomplete or fabricated information, especially with complex cases.
Conclusions:
- ChatGPT-4o's current accuracy and reliability are insufficient for unsupervised clinical use in dental radiography.
- Professional oversight is essential to mitigate diagnostic errors associated with AI interpretation.
- Further AI model refinement and specialized training are necessary for safe integration into dental practice.
