EUVIMEDEUVIMED entdecken
Open Access · CC BY-NC 3.0

Hotspot Evolution and Future Prospects of Large Language Models in Medical Education: A Bibliometric Analysis

Can Huang iD, Wei Liu

Advances in Medical Education and Practice · 2026-07 · Volume 17 · 1-18

DOI: 10.2147/amep.s625578 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advances in Medical Education and Practice, 2026-07. Verfasst von Can Huang und 1 weiteren Autor:innen. Beteiligte Einrichtung: Department of Pharmacy, Beijing You'an Hospital Affiliated to Capital Medical University, Beijing, 100069, People's Republic of China. Die Arbeit wurde bisher 1-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC 3.0).

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Background

Advances in AI and NLP have popularized large language models (LLMs) in medical education. Post-2022 research has proliferated yet remains fragmented; no comprehensive bibliometric mapping systematically outlines this field's global layout, collaboration networks and thematic evolution.

Objective

To clarify publication trends, core contributors, collaboration patterns, research hotspots and evolutionary frontiers of LLMs in medical education via bibliometric analysis, and deliver targeted insights for educational practice.

Methods

This is a systematic bibliometric study. We collected 2016-March 2026 peer-reviewed English papers from WoSCC and Scopus. Metadata were standardized and analyzed with the Bibliometrix R package and VOSviewer to generate publication statistics, collaboration networks, citation metrics and keyword cluster maps.

Results

In total, 1991 papers were included, with an annual growth rate of 59.04%. The US (28.5%) and China (15.9%) led global outputs; Harvard, the University of Toronto and top Chinese scholars dominated contributions, and JMIR Medical Education served as the core journal. Five thematic clusters were identified: education ethics, LLM performance, patient education, clinical reasoning and intelligent assessment. Research priorities shifted from basic neural network exploration to privacy and standardized large-scale LLM deployment after 2024.

Conclusion

LLM medical education research grows rapidly but suffers insufficient empirical evidence, unbalanced themes and delayed governance frameworks. Based on our bibliometric findings, we propose practical optimization schemes: strengthen interdisciplinary research, build matched risk supervision, define LLMs as teaching assistants, and prioritize cohort trials, specialized medical LLMs and unified ethical standards. Academic Contributions: This study provides the latest full-spectrum quantitative mapping of the field, fills gaps in systematic literature review, and offers actionable references for medical educators, curriculum developers and policymakers to realize safe, high-quality LLM integration into medical education.

Schlagwörter

Stichwörter der Autor:innen

artificial intelligencebibliometric analysislarge language modelsmedical educationresearch trends

Autor:innen und Institutionen

  • Department of Pharmacy, Beijing You'an Hospital Affiliated to Capital Medical University, Beijing, 100069, People's Republic of China.

ORCID-Profile: Can Huang

Zitieren

APA

Huang, C., & Liu, W. (2026). Hotspot Evolution and Future Prospects of Large Language Models in Medical Education: A Bibliometric Analysis. Advances in Medical Education and Practice, Volume 17, 1-18. https://doi.org/10.2147/amep.s625578

Vancouver

Huang C, Liu W. Hotspot Evolution and Future Prospects of Large Language Models in Medical Education: A Bibliometric Analysis. Advances in Medical Education and Practice. 2026;Volume 17:1-18. doi:10.2147/amep.s625578

BibTeX
@article{Huang2026,
  author = {Huang, Can and Liu, Wei},
  title = {Hotspot Evolution and Future Prospects of Large Language Models in Medical Education: A Bibliometric Analysis},
  journal = {Advances in Medical Education and Practice},
  year = {2026},
  volume = {Volume 17},
  pages = {1--18},
  doi = {10.2147/amep.s625578}
}
Kostenlos starten →