Generative Artificial Intelligence in Oral Medicine and Radiology: Hype or Hope?

Abstract

Dear Editor, This is in relation to the article titled “Evaluating ChatGPT 3.5 and 4.0 in Oral Medicine and Radiology: A Comparative Query-Based Cross-sectional Study;” published in your esteemed journal.[1] Authors have ably highlighted the potential of generative AI in Oral Medicine and Radiology. However, a little more critical and clear analysis of potential application of ChatGPT in specific aspect of Oral Medicine and Radiology could give some further insight to the readers who wish to explore the same. At the outset, little more clarity in the intent of the study could have been very helpful in understanding whether we are talking about application of ChatGPT in clinical practice or training in Oral Medicine and Radiology. Accuracy, application, practical efficacy, and a clear vision about its exact role must be considered before claiming its application in the specialty. As apparent from the methodology, the study “compared the performance of two versions of ChatGPT in answering a predetermined set of questions.” As per my understanding, this limited the scope of the study in just comparing the efficacy of ChatGPT 3.5 and 4.0 in preparing answers to the academic queries based on standardized prompts. I am unable to understand how exactly this would determine its role in dental education as claimed in the abstract and conclusion. Even in discussion, the claim that the results emphasize the transformative potential of AI in various facets of Oral Medicine and Radiology education remains unsupported factually. A reference has been quoted about study conducted by Gupta et al.,[2] however that study evaluated the accuracy of the two versions of ChatGPT in neuroradiology diagnostics and the scope was well defined. The present study was limited to the answering the academic queries in the subject in that the advanced version, rhetorically; was more efficient due to obvious better technical configuration. It is also unclear how this application exactly helps in the training or education of Oral Medicine and Radiology, unless we are training the students in finding out the answers to the queries using these tools. However, if we are making the users aware of the possibilities of getting inaccurate information based on the diversity of prompts, it shall be duly acknowledged and discussed to increase the suitable and at the same time cautious usage of the generative AI tools. If we speak about applications of AI in dental education solely, several applications have been highlighted by various authors in the past, such as[3,4]: Automated scoring and grading Personalized education Teaching support Generating clinical case scenarios Creating educational content Language conversion In case, we need to explore the application of generative AI in dental education and specifically in training of Oral Medicine and Radiology, we could think of the above aspects. Authors also quote that the study focusses on AI in dental education, stressing its probable use in improving teaching-learning strategies. A little more clarity in deriving this inference would have been much appreciated. We can also not ignore the possibility of unethical usage of generative AI. As far as dental education is concerned, it also seems to have several limitations including questionable data accuracy, bias, over-reliance on the AI tools amongst the students and reduced human engagement. AI seems to be like a boon for the health care sector and has the potential to be so if used appropriately and ethically. On the flipside, it also has the potential to be misused, challenging the academic integrity concerns. As researchers, the onus is on us to highlight both the sides of the story, supported by facts and inferring the results with utmost caution. Dr. Amita Aditya Professor Department of Oral Medicine and Radiology Dr. D. Y. Patil Dental College and Hospital, Dr. D.Y. Patil Vidyapeeth Pimpri, Pune Financial support and sponsorship Nil. Conflicts of interest There are no conflicts of interest.

Ähnliche Arbeiten