Comparing the Performance of Different Artificial Intelligence Tools in Evaluating Dental Morphology Model...

Area of Science:

• Dental Education Technology
• Artificial Intelligence in Medicine
• Morphology Assessment

Background:

• Dental morphology courses require objective evaluation of student-prepared models.
• Traditional assessment methods rely on dental educators' subjective scoring.
• Emerging artificial intelligence (AI) tools offer potential for automated evaluation and feedback.

Purpose of the Study:

• To compare AI model (ChatGPT 5.2, Gemini 3 Pro, Grok 4.1) and dental educator scores for soap tooth models.
• To evaluate the quality of feedback provided by different AI models.
• To assess AI's utility as a supportive tool in dental education assessment.

Main Methods:

• Student-prepared maxillary left canine tooth models were scored by three AI tools and dental educators using a standardized rubric.
• AI-generated feedback quality was assessed by experts using the Global Quality Scale (GQS).
• Statistical analysis included Friedman Test, Bonferroni correction, and Kendall's W coefficient for inter-rater agreement.

Main Results:

• All AI models received significantly higher scores than dental educators (p < 0.001).
• Significant differences in feedback quality were observed among AI models (p < 0.001).
• Gemini 3 Pro provided superior feedback quality compared to ChatGPT 5.2 and Grok 4.1.

Conclusions:

• AI models show promise as supplementary tools for assessment and feedback in dental education.
• Current AI models lack the contextual awareness and personalization to fully replace expert dental educators.
• Further development is needed for AI to match the nuanced feedback of human evaluators.