Developing an AI-powered tool for radiographic feedback on working length determination in pre-clinical endodontic...

Area of Science:

• Artificial Intelligence in Medical Education
• Endodontic Training and Skill Development

Background:

• Accurate working length determination is crucial for root canal treatment success.
• Integrating artificial intelligence (AI) into endodontic training is increasingly important for enhancing dental education.

Purpose of the Study:

• To develop and evaluate a machine learning-based tool for providing feedback on radiographic working length determination.
• To assess the usability and educational support of this AI tool among pre-clinical dental students.

Main Methods:

• A dataset of 3,000 radiographic images was created and labeled for working length determination.
• Twenty-two convolutional neural network (CNN) models were developed and evaluated, with the best integrated into a web platform.
• The tool was piloted with 30 dental students who provided feedback via a Likert-scale questionnaire.

Main Results:

• The AI tool achieved high diagnostic performance, with accuracy ranging from 97%-99% and F1-scores from 95%-98%.
• Students reported positive usability, rating the system highly for clarity, ease of use, and learning support (median scores of 5.0).

Conclusions:

• The AI-powered feedback system is accurate, well-accepted by users, and effectively supports endodontic education.
• This tool offers instant, constructive feedback, beneficial for skill refinement and in large classrooms, with potential for future expansion.