AI is perhaps the technology that offers the greatest hope for more efficient and more individual-based care. At the same time, the ethical aspects contain major challenges. Both perspectives were highly present when DigitalWell Arena hosted the topic at one of Norway's largest AI conferences.
AIM North 2022 is one of Norway's largest AI events. The conference is one of the few that has a pentahelix perspective, which means bringing together the public sector, academia, business, investors and civil society.
During AIM North 2022, the DigitalWell Arena hosted a Break Out session focused on AI in healthcare. The speakers covered a variety of aspects regarding how AI can already contribute in healthcare, but also what challenges need to be overcome in order to benefit from the technology in an ethical and sustainable way.
Important support in the fight against Covid
The speaker who has perhaps made the biggest impression in recent times is Arnoldo Frigessi, professor of statistics at the University of Oslo. During the pandemic, he has been a central figure in Norway's fight against Covid-19. Arnold Frigessi has helped develop statistical prediction models that can predict the development of the spread of infection three weeks ahead. During the pandemic, the algorithm/model has developed to become increasingly precise and accurate, which has meant that healthcare has been better able to adapt its capacity and at the same time given authorities the opportunity to make more well-founded decisions regarding protective measures.
Marie Granander together with Robert Jenssen at AIM North.
Another researcher connected to the University of Oslo, Ishita Barua, also gave a forecast in which areas we can have the greatest hope that AI can contribute to better care. It is about improved workflows, clinical decision support, development of new drugs and monitoring. Last year there were approximately 130 algorithms that have been approved for use in healthcare, to reach further, Ishita Barua highlighted the need to be able to better manage challenges related to ethical issues and risks linked to flaws in the algorithms.
Need for transparency and explanation
Robert Jenssen, director of SFI Visual Intelligence, in this area pressed the need for an XAI – Explainable AI – to make the technology transparent. Among other things, he illustrated the problem with an example of how an AI was trained to identify horses in a picture, where it turned out that the photographer's name was most decisive for the interpretation. A similar challenge arose when an AI was found to be interpreting from which hospital x-rays had been sent, rather than making a medical assessment.
Marie Granander, Process Manager at DigitalWell Arena, moderated the lectures and thought the day offered many interesting perspectives:
- AI, Big Data, Machine Learning and other future-oriented technologies will have a decisive impact on how we cope with future welfare. Precision data that can better diagnose each individual's needs has great potential to improve care, while automated flows and decision support can provide efficiencies. However, the ethical aspect is a key for us to be able to introduce new technology, where understanding and trust are of central importance, says Marie Granander.
All speakers at DigitalWell Arenas
Break Out Session at AIM North:
- Ishita Barua, Oslo University Hospital
- Arnoldo Frigessi, University of Oslo
- Maria-Liisa Bruckert, SQIN
- Robert Jenssen, SFI Visual Intelligence
- Ingrid Vasiliu Feltes, Softthread
- Marie Granander, DigitalWell Arena
