Responsible Machine Learning in Healthcare


An interdisciplinary, invited workshop in Copenhagen on Oct 27th and 28th, 2022.

Find Out More

Aims and Scope


Our aim with this invited workshop is to bring together a diverse and interdisciplinary group of internationally renowned scientists along with junior participants. In lively and critical discussions across disciplinary boundaries, we aim to make substantial progress towards medical machine learning systems that are trustworthy, equitable,
and beneficial for society at large
.

While the workshop is primarily a physical event, a live-stream of the lectures will be made available for free. (See below for details.)

Workshop schedule


The workshop will take place in Festsalen, Øster Voldgade 10.

Program — Thursday, Oct 27th

Time (CET) Program
8:30-9:00Arrival, coffee
9:00-9:10Welcome
9:10-10:40 Enzo Ferrante (Universidad Nacional del Litoral, Argentina)
Fairness of Machine Learning in Medical Image Analysis
[Slides] [Recording]

Fernanda Ribeiro (University of Queensland, Australia)
How fair is your graph? Exploring fairness concerns in neuroimaging studies
[Slides] [Recording]
10:40-11:00Coffee break
11:00-12:30 Damian O. Eke (De Montfort University, United Kingdom)
Data, Power imbalances and AI in Healthcare
[Recording]

Seyed-Ahmad Ahmadi (NVIDIA, Germany)
Towards robust AI in healthcare without jeopardizing patient privacy
[Slides] [Recording]
12:30-14:00 Lunch
14:00-15:30 Thomas Grothe (University of Tübingen, Germany)
Microfoundations of Physician Behavior:
Machine Learning, Predictions, and the (Im)Possibility of Public Health Interventions


Suproteem Sarkar (Harvard University, United States)
A Machine Learning Approach to Breast Cancer Screening
15:30-17:00Poster session & Coffee
18:00-20:30Social dinner

Program — Friday, Oct 28th

Time (CET) Program
8:30-9:00 Arrival, coffee
9:00-10:30 Claes Nøhr Ladefoged (Rigshospitalet, Denmark)
Experience with using AI in clinical routine: Parkinson's disease and synthetic CT
[Recording]

Laure Wynants (Maastricht University, Netherlands)
Fact vs fiction: clinical risk prediction models to support
clinical decision-making in the covid-19 pandemic

[Recording]
10:30-11:00 Coffee break
11:00-12:30 Annika Reinke (German Cancer Research Center, Germany)
(Bench)mark: Pitfalls in AI Validation
[Slides] [Recording]

Enrico Costanza (University College London, United Kingdom)
Human Interaction with ML and AI: User Evaluation Challenges
[Slides] [Recording]
12:30-14:00Lunch
14:00-15:30 Louise Druedahl (University of Copenhagen, Denmark)
Responsible Machine Learning in Health: Regulatory and Legal Aspects
[Slides] [Recording]

David Wong (University of Manchester, United Kingdom)
(What) can healthtech startups teach us about responsible ML?
[Slides] [Recording]
15:30-16:00Closing

Virtual participation


For online participation, please use this Zoom link on both days.

Due to a technical issue, we had to change the Zoom link today (Friday). New link. We apolotize for any inconvenience!

Posters


In addition to the oral presentations, we had a great selection of posters on display during the workshop:

Elena Albu (KU Leuven)
missForest v2 - Missing data imputation for prediction

Timo Freiesleben (Tübingen University)
What Does Explainable AI Explain?

Lu (University of Copenhagen)
Reducing Annotation Need in Self-Explanatory Models for Lung Nodule Diagnosis (Poster, Paper)

Paschal Ochang (De Montfort University Leicester)
Ethical Principles for Responsible Data Governance: Neuroscience Vs AI

Abdolrahman Peimankar (University of Southern Denmark)
An Explainable AI Approach for Lung Cancer Prediction Based on Standard Blood Sample Analysis: A Danish Retrospective Study

Tomer Sagi (Aalborg University)
Teaching Analytics Medical-Data Common Sense (Paper)

Martin Schüßler (TU Berlin)
Do Users Benefit From Interpretable Vision? A User Study, Baseline, And Dataset (Paper)

Raghavendra Selvan (University of Copenhagen)
Carbon footprint of Deep Learning (Paper)

Contact


For any inquiries regarding the workshop, please contact Eike Petersen.

Funding


The workshop is generously funded by the Danish Data Science Academy (DDSA), the Independent Research Fund Denmark (DFF), and the Pioneer Centre for AI.


Code of conduct


We adopt the DDSA Code of Conduct for Events; please follow the link for details. In particular, please take note of the following section, reproduced verbatim from the code of conduct:

Organisers as well as any participant in DDSA activities may contact [the DDSA] about experiences or observations of inappropriate behavior, regardless of whether they wish to place a formal complaint or merely express a concern or seek advice. Please contact the DDSA Named Person, Education & Grant Manager Sniff Andersen Nexø, by calling +45 31 90 35 39, or by sending an email to sniff@ddsa.dk requesting to be contacted.

  • The Named Person will support the person reporting an offensive behavior in deciding if and how to proceed
  • The Named Person has no formal authority but will support the person reporting an offensive behavior in informing the relevant authority if agreed upon
  • The Named Person is not obliged to inform any other authority and is thus able to handle any complaint or concern in full confidentiality
  • Professional consultancy will be included if deemed necessary