Search for a command to run...
Yuan Ren,1 Ying Wang,2 Sheng Dong,1 Wei Su,1 Youtu Wu,1 Shikai Liang,1 Xuejun Yang1 1Department of Neurosurgery, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua University, Beijing, People’s Republic of China; 2Department of Neural Reconstruction, Beijing Neurosurgical Institute, Capital Medical University, Beijing, People’s Republic of ChinaCorrespondence: Xuejun Yang, Department of Neurosurgery, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua University, 168 Litang Road, Changping District, Beijing, 10084, People’s Republic of China, Email Yangxuejunneuro@163.comAbstract: Artificial intelligence (AI) is rapidly transforming medical education through large language models (LLMs), virtual reality (VR), intelligent tutoring systems, and decision-support platforms. These tools enable adaptive instruction, immersive simulation, and real-time feedback, showing strong potential to improve outcomes across health professions training. To explore both opportunities and risks, we conducted a systematic review of PubMed, EMBASE, Web of Science, and Scopus for English-language studies published between January 2015 and May 2025, following the PRISMA framework. Nineteen studies met eligibility criteria. AI modalities identified included LLMs such as ChatGPT, VR-based simulation systems, automated tutoring platforms, and clinical decision-support tools, spanning specialties including radiology, surgery, and psychiatry. Across contexts, AI enhanced examination performance, procedural competence, self-directed learning, engagement, and motivation relative to traditional methods. Students and faculty expressed strong interest and optimism but reported limited formal AI training, favoring interactive practice over didactic lectures. Despite these benefits, concerns consistently emerged regarding algorithmic bias, inaccuracy, data security, and the necessity of human oversight in educational and clinical settings. Ethical issues such as job displacement, the erosion of humanistic care, and the impact on the patient-physician relationship were also highlighted. Limited formal AI training, uneven institutional readiness, and gaps in faculty expertise were common challenges across regions.To harness its transformative potential responsibly, investment is required in faculty development, structured curricula addressing both technical and ethical competencies, and governance frameworks that ensure equitable, transparent, and accountable use. Properly integrated, AI can not only personalize learning and expand access but also support a more inclusive and ethically grounded vision for the future of medical education.Keywords: artificial intelligence, medical education, simulation-based learning, student perceptions, curriculum integration