MICCAI Workshop on Fairness of AI in Medical Imaging (FAIMI)

Dr Andrew King

Biomedical Engineering Dept,

King’s College London

www.kclmmag.org

andrew.king@kcl.ac.uk

Fairness of AI in

Medical Imaging

What does AI fairness mean?

• Fairness is the absence of bias …
• Suppose we have an AI tool that we have trained to diagnose a disease based on an image (sequence)

​

Bias in AI = different levels of performance for different “protected groups” → a form of discrimination

so what is bias?

What is FAIMI?

• FAIMI is an independent academic organisation dedicated to promoting the issue of Fairness of AI in Medical Imaging
• Also a MICCAI Special Interest Group (SIG)
• So far organised 3 free online symposia and 2 MICCAI workshops (third one in 2025)

3

FAIMI web site:�https://faimi-workshop.github.io/

​

MICCAI FAIMI workshop

• First workshop at MICCAI 2023:
• 20 submissions, 19 papers accepted including 4 oral presentations
• Keynote speaker: Judy Gichoya, Emory University USA
• >30 attendees
• Second workshop at MICCAI 2024:
• 17 submissions, 15 papers accepted including 5 oral presentations
• Keynote speaker: Mercy Asiedu, Google Research
• >60 attendees
• Third workshop at MICCAI 2025:
• 26 submissions, 20 papers accepted
• Keynote speaker: Tal Arbel, McGill University Canada

​

4

Workshop scope

• E.g. MICCAI FAIMI 2025 Call for Papers:

5

Why submit to MICCAI FAIMI?

• Showcase your research to a global audience at one of the premier conferences in medical imaging (MICCAI)
• Engage with leading experts in the field, including our organisers, advisory panel and activity committee members
• Contribute to the advancement of ethical and fair AI practices in healthcare
• All accepted papers published in Springer Lecture Notes Computer Science (LNCS) series

6

MICCAI FAIMI awards

• Best oral & best poster awards

​

• Travel awards are available, e.g.
• 2024: 4 registration awards given to authors from low-and-middle-income countries
• 2025: FAIMI HORIZON AWARD - 1 full scholarship covering flights, registration and accommodation
• First author of paper
• Early career researcher
• Attending MICCAI for first time
• From diverse or underserved background

​

7

How to submit to FAIMI

• Prepare an anonymised paper, 8 pages + 2 pages of references
• Use LNCS format - instructions/templates here: https://www.springer.com/gp/computer-science/lncs/conference-proceedings-guidelines
• Submit through CMT: https://cmt3.research.microsoft.com/User/Login
• Dates (from 2025):
• Submission: 24 June
• Decisions: 16 July
• Camera-ready version: 30 July
• Conference: 23 September

8

Stay involved with FAIMI!

• Sign up to the FAIMI newsletter
• Be notified about future events/initiatives, including our (free) online symposia
• Join our Slack community
• Engage with other researchers (and FAIMI organisers) to discuss fairness work and/or ask for help

9