An Artificial Intelligence Solution for Automated Dental Inspection and Charting

Area of Science:

• Artificial Intelligence in Dentistry
• Computer Vision for Medical Applications
• Deep Learning for Healthcare

Background:

• Manual dental charting is time-consuming and error-prone.
• Existing AI solutions for dental charting often rely on radiographs, limiting their use in screening.
• Automating dental charting with intra-oral videos offers a promising alternative.

Purpose of the Study:

• To develop a deep-learning-based system for real-time dental charting using intra-oral videos.
• To establish tooth identification as the initial phase of this automated charting system.
• To evaluate the feasibility of using Convolutional Neural Networks (CNNs) for this task.

Main Methods:

• 100 intra-oral videos were recorded and annotated.
• A YOLOv7x object detection model was developed using a CNN.
• The model's performance was evaluated using mean Average Precision (mAP) at an Intersection over Union (IoU) of 0.5.

Main Results:

• The YOLOv7x model achieved a mAP of 78.8% at 0.5 IoU.
• Precision and Recall for tooth identification were 72.5% and 77.9%, respectively.
• The system demonstrated effective tooth identification from intra-oral video data.

Conclusions:

• This study presents the first use of real-time intra-oral videos for automated dental charting.
• The developed AI approach shows potential for reducing clinical time and costs.
• This technology could improve patient outcomes and access to dental care.