Evaluation of a Novel Veterinary Dental Radiography Artificial Intelligence Software Program.

Area of Science:

• Veterinary medicine
• Artificial intelligence
• Diagnostic imaging

Background:

• Artificial intelligence (AI) is increasingly used in veterinary medicine to aid clinical decisions.
• AI-based software programs are being developed to assist in interpreting veterinary diagnostic images.

Purpose of the Study:

• To evaluate the agreement of a commercially available AI-based software program (AISP) with human evaluators in detecting common radiographic dental pathologies in dogs and cats.
• To assess the sensitivity and specificity of the AISP for specific dental conditions.

Main Methods:

• An AI-based software program (AISP) was used to detect specific dental pathologies (furcation bone loss, periapical lucency, resorptive lesion, retained tooth root, attachment loss, tooth fracture).
• Inter-rater reliability was measured using absolute percent agreement and Gwet's agreement coefficient.
• Sensitivity and specificity were assessed with human evaluators as the reference standard.

Main Results:

• Good to excellent inter-rater reliability was observed between the AISP and human evaluators for detecting specified pathologies.
• The AISP demonstrated a trend of low sensitivity and high specificity.
• A low rate of false positives suggests the AISP could function as a "second set of eyes" for veterinary diagnostics.

Conclusions:

• The AI-based software program shows comparable performance to human evaluators in identifying specific dental pathologies.
• The AISP's high specificity and low false positive rate indicate its potential as an assistive tool, rather than a standalone screening method.
• Understanding the AISP as an aid, not a substitute, may enhance dental radiography utilization and diagnostic accuracy in veterinary practice.