Artificial Intelligence in Medical Education: A citation-based systematic literature review

Abstract

Purpose: This review aims to describe the existing and emerging role of Artificial intelligence (AI) in medical education, as this may help set future directions. Methodology: Articles on AI in medical education describing integration of AI or machine-learning (ML) in undergraduate medical curricula or structured postgraduate residency programs were extracted from SCOPUS database. The paper followed the guidelines of Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) research methodology. Articles describing AI or ML, but not directly related to teaching and training in structured programs were excluded. Results: Of the 1020 documents published till October 15, 2020, 218 articles are included in the final analysis. A sharp increase in the number of published articles was observed 2018 onwards. Articles describing surgical skills training, case-based reasoning, physicians' role in the evolving scenario, and the attitudes of medical students towards AI in radiology were cited frequently. Of the 50 top-cited papers, 16 (32%) were ‘commentary’ articles, 13 (26%) review articles, 13 (26%) articles correlated usefulness of ML and AI with human performance, whereas 8 (16%) assessed the perceptions of students toward the integration of AI in medical practice. Conclusion: AI should be taught in medical curricula to prepare doctors for tomorrow, and at the same time, could be used for teaching, assessment, and providing feedback in various disciplines.