The Comparison of ChatGPT 3.5, Microsoft Bing, and Google Gemini for Diagnosing Cases of Neuro-Ophthalmology
Ruchi Shukla, Ashutosh K Mishra, Nilakshi Banerjee, Archana Verma
Abstract
OBJECTIVE: We aim to compare the capabilities of ChatGPT 3.5, Microsoft Bing, and Google Gemini in handling neuro-ophthalmological case scenarios. METHODS: Ten randomly chosen neuro-ophthalmological cases from a publicly accessible database were used to test the accuracy and suitability of all three models, and the case details were followed by the following query: "What is the most probable diagnosis?" RESULTS: On the basis of the accuracy of diagnosis, all three chat boxes (ChatGPT 3.5, Microsoft Bing, and Google Gemini) gave the correct diagnosis in four (40%) out of 10 cases, whereas in terms of suitability, ChatGPT 3.5, Microsoft Bing, and Google Gemini gave six (60%), five (50%), and five (50%) out of 10 case scenarios, respectively. CONCLUSION: ChatGPT 3.5 performs better than the other two when it comes to handling neuro-ophthalmological case difficulties. These results highlight the potential benefits of developing artificial intelligence (AI) models for improving medical education and ocular diagnostics.