Bibliometric mapping and evolutionary logic of large language models reshaping medical education 2022–2026
TL;DR
A clear knowledge map of existing research in the field of LLMs in medical education is provided and it is suggested that future research should advance toward specialized models for medicine, integration with EHRs and clinical data, and multimodal capabilities.
Abstract
In recent years, Large Language Models (LLMs) have been widely applied in medical education and clinical practice, playing an important role in promoting the digital transformation of medical education. The application of LLMs in the field of medical education has shifted from initial functional validation to technology empowerment, which has enhanced medical teachers` work efficiency, students’ personalized learning, and patients’ self-health management. This review selected core literature from the Web of Science database between 2022 and 2026 and conducted bibliometric and visual analysis using CiteSpace, VOSviewer, and R Bibliometrix. The results indicated that research on LLMs empowering medical education is rapidly growing. The main findings are as follows: (1) BMC Medical Education is the most authoritative journal in this field, and open-access journals such as Healthcare have actively promoted the development of this research area. (2) The key authors in this field include Cheungpasitporn, Thongprayoon, and Klang, with the United States and China ranking highest in publication output. In addition, most core research institutions are located in North America, Europe, and Asia, among which Stanford University is the most influential institution. (3) This review revealed a knowledge system based on model system, scenario applications, academic framework, and reflection, identifies the research areas of seminal literature, and observes an evolutionary process from initial feasibility analysis of LLMs functions to integration with other medical technologies. (4) Current research hotspots mainly focused on core technology, technical platform, and application scenarios, revealing a transition from model validation to teaching and clinical application. Finally, we suggested that future research should advance toward specialized models for medicine, integration with EHRs and clinical data, and multimodal capabilities. This review provided a clear knowledge map of existing research in the field of LLMs in medical education and offers important references for medical teachers and educational policymakers in teaching innovation and clinical exploration.