AI Detects Cavities and Fillings From Toothbrush-Camera Video in Proof-of-Concept Study

Start Discussion
Posted: June 12, 2026

AI Detects Cavities and Fillings From Toothbrush-Camera Video in Proof-of-Concept Study

Edited by Dentaltown staff

A deep-learning system that analyzes video from a toothbrush-mounted intraoral camera detected common dental findings such as decay and fillings with mean average precision above 80%, according to a proof-of-concept study published June 12 in BMC Oral Health.

Researchers built two YOLOv8s detection models using 708 videos recorded with an electronic toothbrush fitted with a miniature intraoral camera. A custom Python script pulled 16,552 frames from the footage, which were filtered to 7,963 frames for a model that identifies dental findings and 3,799 frames for a model that classifies tooth types. Two dental experts annotated decay, fillings, plaque and staining, along with tooth types, using bounding boxes.

The dental-findings model reached an overall mean average precision of 83.0% and an F1 score of 77.9%. It performed best on amalgam fillings, at 94.5%, and decay, at 84.2%.

The tooth-type model posted an overall mean average precision of 88.6% and an F1 score of 86.2%, with its strongest results on molars at 97.7% and premolars at 96.7%.

The models maintained performance across varied footage even without a standardized capture sequence, a point the authors emphasized because video recorded by patients rather than clinicians tends to be inconsistent in angle, lighting and focus.

The study was written by Julia Michelin, Ayesha Nooruddin and Fahad Umer. Umer, the corresponding author, is based at Aga Khan University Hospital in Karachi, Pakistan.

The authors positioned the work as a proof of concept for building AI diagnostics into everyday oral-care devices, with potential applications in remote monitoring, earlier intervention and resource allocation in regions with few dental providers. They cautioned that it remains an early validation effort rather than a clinic-ready diagnostic tool.

Sources:
BMC Oral Health, “Deep learning models for the detection of dental-findings and tooth-types using video data,” by Julia Michelin et al., June 12, 2026 (DOI 10.1186/s12903-026-08861-y): link.springer.com/article/10.1186/s12903-026-08861-y


AI Detects Cavities and Fillings From Toothbrush-Camera Video in Proof-of-Concept Study

Views: 15
Start Discussion
Dentaltown Magazine
June 2026
Dentaltown Magazine June 2026 Read Current Issue Read Archived Issues
Sponsors
Townie Perks
Townie® Poll
What part of a dental office do you feel makes the strongest first impression on patients?
  
Site Menu
Message Boards Magazine Continuing Education Podcasts Blogs Classifieds Dental News Events Townie® Poll Ebooks Townie Choice Awards
Company Info
Contact Us Advertise With Us Howard Farran DDS, MBA Policies and Agreements
Site Help
The Dentaltown Team, Farran Media Support
Phone: +1-480-445-9710
Email: support@dentaltown.com
©2026 Dentaltown, a division of Farran Media • All Rights Reserved
9633 S. 48th Street Suite 200 • Phoenix, AZ 85044 • Phone:+1-480-598-0001 • Fax:+1-480-598-3450