Table 1 Potential curricular and extracurricular learning opportunities for artificial intelligence in medicine.

From: What do medical students actually need to know about artificial intelligence?

Curricular objectiveDelivery recommendationsExtracurricular objectiveDelivery recommendations
Promote physicians to be data-savvy consumers
Students should be able to critically evaluate AI claims and understand the connection between models and clinical realities.
Actively engage students with hands-on workshops focused around:
Recognizing appropriate potential applications of AI to health data
Understanding how to discern between different methods that can be applied to data (e.g. the distinction between prediction and causal inference approaches)
Promote student interest groups in AI
Interested students should be encouraged to connect and build networks around their shared AI focus.
Extend the broadly-used format of the “student interest group” to AI, enabling students to organize and autonomously host initiatives such as:
Seminar events with prominent AI in Medicine speakers
Hackathons and datathons in collaboration with computer science and engineering students
Instill durable fundamental concepts about AI, while avoiding technical specifics
It is more important for students to have a robust conceptual understanding of AI and the structure of clinical data science than to understand constantly changing technical specifics.
Incorporate lecture and self-learning module content around:
The basic pipeline of data acquisition, cleaning, analysis, and visualization
Issues with data stewardship and data quality assurance in healthcare.
Classes of machine learning approaches and common issues with design and integration of AI into clinical practice
Facilitate connections between medical students and industry in the health-AI space
As AI in medicine is not siloed to solely academia or industry, students should have the opportunity to be exposed to the AI ecosystem in their local and broader communities
Leverage partnerships (either at the Faculty level or the student group level) to offer:
Student site visits to start-ups to learn about entrepreneurship and the creation of health AI products and services
Student research opportunities with health AI companies, or public-private partnerships
Introduce frameworks for approaching ethical considerations, both clinically and at a systems level
Students should appreciate fairness, accountability, and transparency as core AI analogues to the traditional bioethics principles of beneficence, non-maleficence, autonomy, and justice15.
Students should participate in interactive case-based workshops and seminars lead by AI and ethics experts focused on:
The special considerations AI requires at clinical and system levels in a case-based format
How fairness, accountability, and transparency directly relate to core clinical values of beneficence, non-maleficence, autonomy, and justice which must permeate through all aspects of their care
Provide longitudinal programs to give students hands-on experience with real-world AI projects
Theoretical knowledge should be supplemented with practical, real-world experience through formalized programs.
Longitudinal programs can include but are not limited to:
“Computing for Medicine”, a validated 14-week course offered to preclinical students in UofT’s MD program to promote computer literacy, algorithmic thinking, and cross-domain collaboration9
Non-technical projects involving AI or data science, such as using design-thinking approaches to implement existing AI tools into clinical practice and workflows
Promote computer science/data science as a dual-training path for MD/PhD and MD/MSc students
Students should be provided with partnered, formalized learning opportunities that provide training at the intersection of health and data science.
Establish partnerships with institutes across computer science, biomedical engineering, the basic sciences, and public health, such as:
U of T Faculty of Medicine’s partnerships with the Vector Institute for Artificial Intelligence, and Schwartz Reisman Institute for Technology and Society
Harvard Medical School’s Collaborative Health Sciences and Technology MD/PhD Offerings with MIT
Encourage cross-disciplinary collaborations between medical students and data scientists
Students should build interdisciplinary networks, and be encouraged to connect and collaborate with peers across faculties
Take active steps to break past disciplinary silos through initiatives such as:
Shared “AI in Medicine” journal clubs open to both medical students and computer science/engineering students
Collaborative events such as “datathons”, wherein ad-hoc interdisciplinary teams compete to answer clinical questions on open database