Performance Evaluation of ChatGPT in Forensic Medicine (Including Medical Jurisprudence & Toxicology) of Third Professional MBBS: Part 1

Abstract

Background: ChatGPT, the state-of-the-art innovation in the field of Artificial Intelligence (AI) is gaining massivepopularity with its notable ability to generate human-like responses within the shortest period. The utility ofChatGPT is being assessed possibly in all fields now with medical education predicted to be one of the key areasto be highly benefitted by ChatGPT.Objective: We aim to evaluate the performance of ChatGPT in Forensic Medicine of the third professional MBBSpart 1.Methods: The performance of ChatGPT was assessed using the final university examination question papers. Thequestions were used to converse with ChatGPT and each of the responses was recorded. The answers generatedby ChatGPT were later evaluated by the expert faculties in the Department of Forensic Medicine. Results: All ourquestions regarding the subject were satisfactorily addressed by ChatGPT. The answers were concise without anypictorial representation. Before generating content related to issues like “sodomy”, potential content violationmessages were sent.Conclusion: ChatGPT has performed moderately in the evaluation. Although there were invalid and incorrectresponses, with additional inputs and training, these chatbot systems can be improved to assist medical educatorsin various subjects and can emerge as one of the innovative teaching-learning strategies in medical education. Wecan successfully use these technologies to improve medical education and, ultimately, patient care by leveragingAI’s benefits while addressing its drawbacks.