فا Sign in

AI in Medical Education and Research

Practical uses of AI tools for teachers, students and researchers, together with their limits and the principles of responsible use.

Generative AI can give wrong answers that sound convincing. Never use its output without expert review, especially on clinical topics. This site gives no clinical advice; never enter identifiable or clinical patient data into public tools.

AI in medical education

Common uses for teachers

Lesson planning

Draft objectives, activities and timing that you refine and approve.

Question writing and review

Suggest multiple-choice questions from lesson text and review distractors. Check scientific accuracy yourself.

Simulation scenarios

Build fully fictional patient scenarios for clinical reasoning practice. Do not use real patient data.

Feedback and summarising

Draft feedback on assignments and summarise long texts; the final version is yours.

Personalised study

Students can ask for an explanation or practice questions, as long as course rules allow it.

Translation and rewriting

Simplify text for different audiences and make first-pass translations; always review technical terms.

A good prompt template

Five parts: role, goal, audience, output format and constraints. Replace the values in [ ].

You are an experienced [field] teacher.
Goal: design a 45-minute session on [topic].
Audience: [level/year] students who already know [prerequisite].
Format: a timetable with columns for activity, duration and assessment method.
Constraints: at most 3 measurable objectives; fictional scenarios only; flag every scientific claim as 'needs source check'.

AI in medical research

A research assistant, not a replacement for the researcher

Searching and literature review

Brainstorm keywords and summarise papers. Verify every reference in trusted databases; tools sometimes invent sources.

Scientific writing and editing

Language editing, structuring and tightening. Authors remain responsible for content and claims.

Data analysis and coding

Explain code errors and suggest analysis methods. Confirm results independently or with a statistician.

Limitations

Errors and bias, outdated knowledge, and weak methodological judgement. Do not paste unpublished or sensitive data into public tools.

Transparency in publishing

Most journals and publishing guidance (e.g. ICMJE) state AI cannot be an author and its use must be disclosed. Read the target journal’s policy before submitting.

Reproducibility

Record the tool, version and key prompts so the work can be traced.

Machine learning and data analysis

Coming soon; a dedicated section is being planned

Planned topics

  • ML fundamentals for medicine
  • Statistics and data analysis with Python, R and SPSS
  • Model evaluation and interpreting results
  • Hands-on projects and showcases

Get notified at launch

Leave your email; you will join the site newsletter and we will tell you when this section launches.

Ethics and usage policy

Suggested principles; university and journal rules always take precedence

  1. Patient privacy
    Never enter identifiable or clinical data into public tools; anonymise fully and obtain ethical approval when needed.
  2. Accuracy and review
    Check every output against trusted sources and expert judgement. Final responsibility stays with the human.
  3. Transparency
    Disclose AI use in research, publication and coursework as the relevant rules require.
  4. Academic integrity
    Students follow each course’s AI rules; presenting tool-generated work as one’s own is misconduct.
  5. Fairness and bias
    Tools may be biased (gender, ethnicity, age). Be especially careful in decisions that affect people.