JMIR Publications

ABSTRACT

Background:

In the 21st century, Generative Artificial Intelligence (GAI) drives medical education towards enhanced intelligence, personalization, and interactivity. With its vast generative abilities and diverse applications, GAI redefines how educational resources are accessed, teaching methods are implemented, and assessments are conducted.

Objective:

This study reviewed the current applications of GAI in medical education, analyzed its opportunities and challenges, and identified its strengths and potential issues in educational methods, assessments, and resources. We aimed to capture GAI’s rapid evolution and multidimensional applications in medical education, thereby providing a theoretical foundation for future practice.

Methods:

This scoping review used PubMed, Web of Science, and Scopus to analyze literature from 2023 to October 2024, focusing on GAI applications in medical education. Following PRISMA-ScR guidelines, 5,991 articles were initially retrieved, with 1,304 duplicates removed. Two-stage screening (title/abstract and full-text review) excluded 4,543 articles, yielding 137 studies for final synthesis. We included (1) studies addressing GAI’s applications, challenges, or future directions in medical education, (2) empirical research or case analyses, and (3) English-language articles. We excluded systematic reviews, meta-analyses, commentaries, non-GAI technologies and studies centered on other fields of medical education (e.g., nursing or family medicine). We integrated quantitative analysis of publication trends and Human Development Index (HDI) with thematic analysis of applications, technical limitations, and ethical implications.

Results:

Analysis of 137 articles revealed that 77% originated from countries or regions with very high HDI, with the United States contributing the most (35 articles); 12% from high HDI countries, 6% from medium HDI countries, and 1% from low HDI countries, with 4% involving cross-HDI collaborations. ChatGPT was the most studied GAI model (122 articles), followed by Gemini (21 articles), Copilot (10 articles), Claude (6 articles), and LLaMA (4 articles). Thematic analysis indicated that the application of GAI in medical education mainly embodies the diversification of educational methods, the scientific evaluation of educational assessments, and the dynamic optimization of educational resources. However, they also highlighted current limitations and potential future challenges, including insufficient scene adaptability, data quality and information bias, over-reliance and ethical controversy.

Conclusions:

The application of GAI in medical education exhibits significant regional disparities in development, and statistical findings from model research reflect that researchers have certain preferences in its usage. GAI holds immense potential for transforming medical education, but widespread adoption requires overcoming complex technical and ethical challenges. Grounded in the theory of symbiotic agency, we advocate for developing specialized models and constructing an integrated system based on general large language models for specialized model incorporation, improving model adaptability and localization, promoting resource sharing, refining ethical governance, and building an educational ecosystem that fosters human-machine symbiosis, enabling deep integration of technology and humanism and advancing medical education towards greater efficiency and human-centeredness.