Deep learning for automated dental plaque index assessment: validation against expert evaluations

Area of Science:

• Artificial Intelligence in Dentistry
• Digital Dentistry Innovations
• Oral Hygiene Assessment Technologies

Background:

• Artificial intelligence (AI) is advancing healthcare, particularly in diagnostics and clinical decision-making.
• Automated dental plaque detection can enhance patient care and remote oral hygiene assessments.

Purpose of the Study:

• Develop and evaluate a deep learning (DL) system for automatic dental plaque detection and quantification from intraoral images.
• Utilize a standardized plaque index for accurate AI model training and validation.

Main Methods:

• Collected intraoral images from 70 participants before and after applying a plaque-disclosing agent.
• Labeled tooth data using LabelMe software for DL model training and validation.
• Compared the DL system's performance against assessments from an experienced dentist and a dental hygienist.

Main Results:

• The DL model achieved a micro-average accuracy of 73.67% and macro-average accuracy of 65.15%.
• The system demonstrated a precision of 76.34%, recall of 65.15%, and an F1 score of 66.15%.
• Statistical analysis revealed no significant difference between the DL model and the experienced dentist's performance (P > 0.05).

Conclusions:

• The DL-based system effectively automates dental plaque evaluation from images, matching clinician performance.
• AI-driven plaque assessment tools hold significant potential for digital dentistry and remote patient monitoring.
• This technology could enhance oral healthcare delivery and patient outcomes.