“AI is something that affects us all”

“What I am most looking forward to is hearing about all the AI research conducted at Uppsala University and the impact it might have,” says Åsa Johansson, Senior Lecturer at the Department of Immunology, Genetics and Pathology. Photo Mikael Wallerstedt.
At the Olof Rudbeck Day, 18 October, the theme is “AI for health and care – opportunities and risks”. This day offers the public a chance to learn more about a subject that will become increasingly important in the future.
A multitude of experts from both Uppsala University and other higher education institutions will discuss AI in healthcare from various perspectives, including clinical, preclinical, legal, and ethical viewpoints.
"What I am most looking forward to is hearing about all the AI research conducted at Uppsala University and the impact it might have. AI research is both complex and perceived as such, so we want to offer a day full of research explained in a simple and understandable way by numerous experts," says Åsa Johansson.
Åsa is a Senior lecturer and researcher at the Department of Immunology, Genetics, and Pathology at Uppsala University, as well as one of the two programme coordinators for this year's Olof Rudbeck Day.
"This year, we have secured many excellent speakers, both internally from Uppsala University and externally. We felt the need to structure it a bit, so we have divided the day into three parts: preclinical aspects of AI, clinical aspects of AI, and ethical and legal aspects of AI. Each part will conclude with a panel discussion where attendees will have the opportunity to ask questions," says Åsa.
AI, a significant part of society
AI has already become a major part of our daily lives and has been with us for many years, even if we do not always think about it. However, in healthcare, the major breakthrough is still in the future.
"AI is something that affects us all today. Whether we like it, are sceptical of it, or even fear AI, we have a relationship with it – sometimes perhaps without even knowing or thinking about it. I sometimes feel that we tend to overinterpret the potential of AI," says Åsa, continuing:
"Many believe that AI will be able to replace everything in society. Although AI is increasingly used in preclinical research, it is important to map out how AI can be used as a tool in healthcare. How can it actually be used? In which areas? And not least, who is responsible for the AI if it makes a mistake?"
The questions Åsa raises are likely the same questions many others also have. During the Olof Rudbeck Day, the ethical and legal aspects will also be discussed.
"I believe that this part of the day will see the most questions from the public, the questions that each of us ponders."
At the beginning of the AI journey
As a researcher, Åsa Johansson has focused on identifying various molecular and lifestyle-related patterns in individuals and how these affect disease risk. This is to ultimately use different molecular or lifestyle profiles to understand each individual's disease risk.
"At this stage, I see that AI could be helpful. Identifying patterns in medical history, molecular markers, and lifestyle factors involves handling enormous amounts of data. It is in this work that we see potential for an AI model within our research," she explains, elaborating:
"We are at the beginning of our AI journey, one might say. It all comes back to the uncertainty of what AI does, can, and cannot do. When, like in my research group, we look for risk factors, we are interested in what we call causal factors – that is, seeing actual causal relationships between different events. When we fall ill, hundreds of molecules are released, and we do not know if they actually caused the disease or if they are secreted to protect the body and fight the disease. That is where we are and trying to discuss how AI could help and how much we can trust it."
A day for questions and answers
Like most people, Åsa Johansson does not see AI as black or white. There are both opportunities and challenges, but a crucial part is making AI reliable.
"In my research, it is the enormous amounts of data that AI could analyse to identify and estimate disease risks. But it is one thing to get something to work in a study and another for it to work in reality within healthcare. What has been seen, for example, is that AI works very well in medical imaging, for instance, to detect tumours in mammography images – in some cases better than an expert performing a manual assessment. My hope is that the Olof Rudbeck Day will be an opportunity to clarify and discuss both opportunities and challenges, so those attending the lectures can get better answers to questions they may be uncertain about," concludes Åsa Johansson.
Robin Widing