PERFORMANCE OF CHATGPT 4.0 ON OPHTHALMOLOGY BOARD EXAM PRACTICE QUESTIONS

To evaluate the performance of ChatGPT 4.0 in answering ophthalmology questions from two popular board preperation question resources

İzmir Democracy University, Buca Seyfi Demirsoy Training and Research Hospital Department of Ophthalmology, İzmir, Turkey.

Of 398 available questions in https://ebodtraining.com/ question bank 345 (86.6%), text-based questions were selected and organized in order to provide 35 questions in each subspecialty. Questions from https://www.eyedocs.co.uk/ were selected randomly in order to provide the same amount in each subspecialty and the other resource. Responses and percantages of correct answers were recorded in addition to the performances regarding to the subspecialties.

In https://ebodtraining.com/, ChatGpt 4.0 answered 81.4% options correctly. ChatGPT-4.0 was detected weakest in ‘Neuroopthalmology’ and ‘Glaucoma, cataract and refractive surgery’ sections (68.57% and 69.71% correct answer rates respectively.) The best performances were generated in ‘General medicine relevant to ophthalmology’ and ‘Ophthalmic pathology, microbiology’ sections. (95.42%, 91.42% correct answer rates respectively.) In https://www.eyedocs.co.uk/, ChatGpt 4.0 answered 67.53% questions correctly. The weakest performance was in ‘Glaucoma and cataract’ and ‘External eye and cornea’ sections (54.28% and 57.14% correct answer rates respectively) and the best was received from ‘Pathology and microbiology’ section (94.28% correct answer rate).

ChatGPT can be a valuable tool in getting prepared for board examinations with its unique practice opportunities for students. In modern world, time is getting more precious every day, and with the help of AI models, students can receive information and explanation rapidly. In addition, with the advantage of asking further questions, students can access personalised answers which can reduce time consumption and results in tailored learning experience.