Related Experiment Video
Updated: Jul 17, 2026
Guided Endodontics: Three-Dimensional Planning and Template-Aided Preparation of Endodontic Access Cavities
Published on: May 24, 2022
A comparison of endodontists and artificial intelligence chatbots in regenerative endodontic procedures: knowledge
Güzide P Sezgin1, Hatice Sağlam2, Sema Sönmez Kaplan1
1Department of Endodontics, Faculty of Dentistry, Biruni University.
Purpose:
This study aimed to compare the knowledge and reliability of endodontists and artificial intelligence (AI) chatbots (ChatGPT 4o and DeepSeek-R1) in relation to regenerative endodontic procedures (REPs).
Methods:
A 19-item survey addressing case selection, materials, complications, success criteria, and follow-up in REPs was administered to endodontists and AI chatbots. Four clinical case scenarios with sub-questions required participants to rank treatment options by preference. Responses were analyzed for knowledge and reliability using Kruskal-Wallis, chi-square, and Fisher's exact tests; Cronbach's alpha assessed internal consistency.
Results:
ChatGPT 4o demonstrated a significantly higher knowledge rate for REPs (100%) than endodontists (61.5%) and DeepSeek-R1 (56.9%) (P < 0.05). It also achieved the greatest reliability (Cronbach's alpha = 0.852). In the first three case scenarios, ChatGPT-4o most frequently favored REPs, while in the fourth scenario, apexification or conventional root canal therapy was more often preferred.
Conclusion:
ChatGPT 4o exhibited superior knowledge of REPs and reliability compared to endodontists and DeepSeek-R1, which showed moderate and low reliability, respectively. ChatGPT 4o consistently favored REPs for managing immature permanent teeth. As AI use expands in healthcare, optimizing its integration into clinical endodontic decision-making is essential to maximize patient benefit.
