Artificial Intelligence Education for Medical Students: A Systematic Review


  • Aryan Ghaffarizadeh University of British Columbia
  • Nikola Pupic University of British Columbia
  • Ricky Hu Queen's University
  • Rohit Singla University of British Columbia
  • Kathryn Darras University of British Columbia
  • Anna Karwowska
  • Bruce Forster University of British Columbia


artificial intelligence, curriculum development, undergraduate medical education


Introduction: In the recent literature, several groups have advocated for the incorporation of artificial intelligence (AI) training and literacy into medical curricula to prepare future physicians for  its  use. [1- 3] Currently, there is no uniform curriculum incorporated into medical training, which may be ascribed to obstacles such as the selection of topics with the proper breadth and depth. This systematic review seeks to identify and compile the existing evidence-based recommendations as key steps towards an AI curriculum in undergraduate medical education.

Methods: MEDLINE, EMBASE, CINAHL, ERIC, NCBI, and Web of Science were searched from database inception to May 2022 for articles addressing AI education in undergraduate medical education (UGME). The search terms of “medical education”, “artificial intelligence”, “medical curriculum”, and “medical program” were used and combined with boolean operators “AND”, “OR”, and “ADJACENT”. The inclusion and exclusion criteria for this study were determined a priori; studies about UGME with fair quality or higher using the Newcastle-Ottawa scale were included. A thematic analysis was performed to identify core themes.

Results: The original search yielded 991 studies after duplicates were removed. After title, abstract, full-text screening, and reference mining, 38 studies were included for analysis. The studies were separated into two categories: survey (n = 18) and interventional (n = 21). A thematic analysis identified six themes: ethics (n = 11, 28.9%), theory and application (n = 15, 39.5%), communication (n = 11, 28.9%), collaboration (n = 7, 18.4%), quality improvement (n = 9, 23.6%), and perception and attitude (n=3, 7.9%). Within ethics, subthemes of patient and data ethics emerged. Theory and application was further divided into knowledge needed for practice and for development. Communication was stratified as being for clinical decision-making, for implementation, and for knowledge dissemination.

Conclusion: Overall, the six identified themes could serve as a useful framework in building a comprehensive AI curriculum for UGME. Future work on the implementation and integration of the themes into UGME curricula is required.

Author Biographies

Aryan Ghaffarizadeh, University of British Columbia

Medical Student - Year 2

Nikola Pupic, University of British Columbia

Medical Student - Year 2

Ricky Hu, Queen's University

Medical Student - Year 4

Rohit Singla, University of British Columbia

Medical and Doctor of Philosophy Student - Year 5

Kathryn Darras, University of British Columbia

Radiologist & Nuclear Medicine Physician, Vancouver General Hospital
Clinical Assistant Professor, University of British Columbia
Director, Undergraduate Radiology Education, University of British Columbia
Co-Program Director, Nuclear Medicine Residency Program, University of British Columbia

Anna Karwowska

Dr. Anna Karwowska is the Vice-President, Education, for the Association of Faculties of Medicine of Canada (AFMC) and an Associate Professor in the Department of Pediatrics, Faculty of Medicine, University of Ottawa.

Bruce Forster, University of British Columbia

Dept of Radiology
UBC Faculty of Medicine
Member, IOC Medical and Scientific Games Group
Co-Chair, National Artificial Intelligence Council for Healthcare




How to Cite

A. . Ghaffarizadeh, “Artificial Intelligence Education for Medical Students: A Systematic Review”, CMBES Proc., vol. 45, May 2023.