Innovating dental diagnostics: ChatGPT's accuracy on diagnostic challenges

Authors:

Abstract
Introduction: Complex patient diagnoses in dentistry require a multifaceted approach which combines interpretations of clinical observations with an in‐depth understanding of patient history and presenting problems. The present study aims to elucidate the implications of ChatGPT (OpenAI) as a comprehensive diagnostic tool in the dental clinic through examining the chatbot's diagnostic performance on challenging patient cases retrieved from the literature.Methods: Our study subjected ChatGPT3.5 and ChatGPT4 to descriptions of patient cases for diagnostic challenges retrieved from the literature. Sample means were compared using a two‐tailed t‐test, while sample proportions were compared using a two‐tailed χ2 test. A p‐value below the threshold of 0.05 was deemed statistically significant.Results: When prompted to generate their own differential diagnoses, ChatGPT3.5 and ChatGPT4 achieved a diagnostic accuracy of 40% and 62%, respectively. When basing their diagnostic processes on a differential diagnosis retrieved from the literature, ChatGPT3.5 and ChatGPT4 achieved a diagnostic accuracy of 70% and 80%, respectively.Conclusion: ChatGPT displays an impressive capacity to correctly diagnose complex diagnostic challenges in the field of dentistry. Our study paints a promising potential for the chatbot to 1 day serve as a comprehensive diagnostic tool in the dental clinic.

This publication has 11 references indexed in Scilit: