Evaluation of Artificial Intelligence Conversational Models in Providing Information on Dental Implants: A Comparative Analysis of ChatGPT, Gemini and MedGebra.

AIMS AND

OBJECTIVES

This cross-sectional study aims to compare the accuracy, precision and clarity of responses to dental implant questions from three artificial intelligence (AI) models: ChatGPT (OpenAI, GPT-4), Gemini (Google) and MedGebra.

MATERIALS AND

METHODS

56 questions from an EAO patient guide were used and responses from each AI system were evaluated independently by two oral surgery specialists using a four-point Likert scale for correctness, precision and clarity.

RESULTS

ChatGPT consistently scored highest especially in clarity with perfect marks from both evaluators. Gemini ranked lower in accuracy and precision but exhibited similar clarity. MedGebra received the lowest ratings on all criteria. Though consistency generally remained excellent, the inter-rater reliability study revealed some variance in agreement levels between two examiners.

CONCLUSION

ChatGPT and Gemini provided more accurate and informative responses on dental implants questions compared to MedGebra. According to the findings of the present study AI chatbots have potential for enhancing patient education on dental implants but professional oversight is crucial for ensuring accuracy and reliability. The study highlights the promising and potential benefits of large language models (LLMs) in dental patient education.