Comparative analysis of GPT-3.5 and GPT-4.0 in Taiwan's medical technologist certification: A study in artificial intelligence advancements.

Abstract

This study examines the comparative effectiveness of GPT-3.5 and GPT-4.0, in the certification of medical technologists (MT) in Taiwan, exploring their adeptness in processing complex medical language and their contributory role in the educational and communicative aspects of professional healthcare training. This study used GPT-3.5 and GPT-4.0 to test the medical laboratory technician professional college entrance examination questions. The questions in different fields, including six subjects, such as Clinical Physiology and Pathology, Hematology, and Blood Bank, among others were answered one-on-one using two generative pretrained transformer (GPT) versions, simulating the situations during exam preparation. A total of 480 questions were analyzed and the results showed that both versions of the GPT met the certification standards. Version 4.0 was better than version 3.5 for all subjects, particularly in Clinical Biochemistry (score = 96.25) and Microbiology (score = 91.25). Outstanding performance compared to version 3.5, which had an average score of 65.42 and a maximum score of 77.5. Overall, version 4.0, which was significantly better than version 3.5 in both median and average scores, reflects a significant improvement in professional knowledge processing capabilities. The GPT can provide valuable support for both the upstream and downstream processes of MT certification. Future research can further explore the application of GPT in different educational and certification contexts and improve the passing rate of medical personnel in the certification process. This study provides useful information for exploring the potential applications of GPT in certifying medical examiners. Furthermore, it provides new directions for future research in medical education.