Scholarlink

<p>Artificial intelligence (AI) has demonstrated clinical potential, yet its influence on medical education remains limited. This review explores AI applications in medical education, evaluates available evidence and considers future applications. We conducted a scoping review (PubMed, MEDLINE, SCOPUS, Google Scholar; 2010–2022) identifying 42 relevant peer-reviewed articles. Four key themes emerged: surgical skills assessment, radiology training, interactive learning, and text interpretation. Current applications enhance surgical simulation and facilitate interactive learning. These tools may evolve towards comprehensive and individualised educational aids. Despite promising early applications, evidence on educational and clinical outcomes remains limited. Future research should prioritise validated outcomes in larger trials to confirm generalisability and address AI limitations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Artificial Intelligence in Medical Education: a Scoping Review of the Evidence for Efficacy and Future Directions

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年4月2日

查看全文
  • Kody Shaw,
  • Marcus A. Henning,
  • Craig S. Webster

摘要

Artificial intelligence (AI) has demonstrated clinical potential, yet its influence on medical education remains limited. This review explores AI applications in medical education, evaluates available evidence and considers future applications. We conducted a scoping review (PubMed, MEDLINE, SCOPUS, Google Scholar; 2010–2022) identifying 42 relevant peer-reviewed articles. Four key themes emerged: surgical skills assessment, radiology training, interactive learning, and text interpretation. Current applications enhance surgical simulation and facilitate interactive learning. These tools may evolve towards comprehensive and individualised educational aids. Despite promising early applications, evidence on educational and clinical outcomes remains limited. Future research should prioritise validated outcomes in larger trials to confirm generalisability and address AI limitations.