Lund University is a member of LERU

(League of European Research Universites)

LERU Artificial Intelligence

Recently, Artificial Intelligence (AI) was identified as a topic of great societal importance and with a scope that may overarch several LERU groups. AI was recently scheduled on the agenda of several LERU meetings, covering different perspectives: policy-wise, content-wise, programme call oriented. Various contacts between members from LERU universities and LERU Policy Officers illustrated a general need for structured information on EU policy and programmes and clarification how to deal with the challenges and opportunities regarding AI in an academic context.

To capitalize on these initiatives and needs, LERU initially scheduled the joint BIOM, NATU, SSHU meeting in Leuven and the RESE meeting in UCL. However, it was clear that a more structured and targeted approach needed to be set up, hence the Townhall meeting.

Virtual AI Townhall meeting 30 October on

AI-möte September 2020. mLERU AI in Health Sciences Meeting
Notes from an online meeting on AI in the health sciences
Who should learn it?

  • Everyone, but at different levels
  • Medical students who do not end up as a medical doctor can also benefit from it as awareness of AI and related issues is v. important in (e.g. policy) roles too.

What should they learn?

  • Basic, broad ideas may be more useful as they need to be educated as critical thinkers about AI and to be able to understand the topic, without necessarily being able to programme.
  • Programming may demoralise them, and concepts change all the time so they would quickly become out of date in terms of their knowledge.
  • AI is separate from other computer sciences (i.e., automatic image processing, algorithmics, decision support systems). It is not machine learning. A lack of understanding now means that they are often used interchangeably.

When should they learn it?

  • Try to integrate it into existing medical curriculum wherever possible whilst recognising it is already very full.

Why should they learn about AI?

  • Important to learn about the disadvantages and risks, i.e., how AI on 3D imaging may alter the result and how biases against certain groups (e.g. women, people of colour) can be inadvertently built into the AI algorithm.
  • Teaching now will help protect them from fake news/hype surrounding AI and help them keep an open mind about it/its results.

How should they learn it?

  • From multiple teachers, lawyers, and computer scientists in a course coordinated by a medical expert so that everyone stays close to their expertise, but the medical expert gives the overarching medical viewpoint.
  • For the paramedical disciplines, training the teachers and international collaboration approaches could be valuable as there are not many institutions who have teachers with AI skills ready in all (para)medical disciplines. You need to focus on teaching the teacher as technology like AI moves so quickly.

What can we use it for?

We can use it in robotic surgery – Barcelona are collaborating with J&J and Google to analyse thousands of surgeries. The focus is on robotic surgery. The project uses thousands of surgeries from many different surgeons in standardized procedures. The fact the same steps are used throughout each surgery enables the system to compare the skill level of the surgeon concerned

History of the september meet:

  • Plan initiated in January 2020 for a 2-day summer workshop in Leiden
  • LERU-aproved in March
  • Moved to shorter online event 4 September due to CoVID19
  • Draft a position paper with recommendations, as input for further discussions

Primary workshop objectives
AI: What exactly should be taught when in the curriculum and by whom
1) during the Medical Degree curricula
2) during general health sciences and biomedical research curricula
Secondary workshop objectives
Identify cross-cutting topics, e.g.:

  • Legal and ethical framework for AI in health sciences.
  • AI as a health science teaching tool
  • Public-private developments for AI teaching in the health science domain
  • Teaching the teachers: optimal profile of a Health-Science AI teacher?
  •  Benchmark AI added value in health sciences