COLLABORATION. AI Innovation of Sweden, recently opened in Gothenburg, is one of Sweden’s largest investments in artificial intelligence (AI). The initiative is being funded by Vinnova (Sweden’s innovation agency, and Region Västra Götaland along with a large number of partners, including the University of Gothenburg and Chalmers University of Technology.
AI Innovation of Sweden has its headquarters at Lindholmen Science Park, where the official opening took place on February 6. The initiative aims to create a leading environment for collaboration on artificial intelligence. More than 40 partners from industry, academia and the public sector–including Chalmers University of Technology, the University of Gothenburg and Region Västra Götaland (RVG)–have joined the initiative so far. As a national center of collaboration, it will accelerate innovation and research in AI with a focus on practical applications. Interested parties within the center can meet in Gothenburg or at co-location sites in Stockholm and Skåne.
“The new AI Innovation of Sweden center makes it possible for the private sector and organizations to participate in developing the field and improving their expertise within AI,” said Anders Carlberg, head of R&D and regional development in RVG, in connection with the opening. If the center succeeds in involving others, it can have a major impact on Sweden’s future effectiveness and competitiveness.”
The opening ceremony also unveiled the online training course “Elements of Artificial Intelligence,” which teaches the basics of AI. The course was created by the University of Helsinki.
“In Finland one percent of the population has already taken the course. It is free and open to everyone, and I urge all who are interested in artificial intelligence to enroll in the online course,” says Henrik Hagberg, Sahlgrenska Academy pro-dean and a member of the steering group for AI Innovation of Sweden.
You can find “The Elements of Artificial Intelligence” course here: https://www.elementsofai.com/.
SCAPIS is first up
Life Science is one of four major research domains included in the AI initiative. The other three domains are society, technology and finance.
Hagberg says, “The initiative originated with Lindholmen Science Park, which worked hard to create a structure that gives us a joint ‘Data Factory’ where we can submit large quantities of information for data processing.”
One of the first research projects to be carried out at AI Innovation of Sweden is a sub-project within SCAPIS, the unique research study led by Sahlgrenska University Hospital and the University of Gothenburg, which has collected enormous amounts of data on more than 30,000 individuals.
“SCAPIS is a unique study that has collected large amounts of data, which we believe will work very well as a pilot project within Life Sciences for AI Innovation of Sweden,” says Hagberg. “SCAPIS is first up, but in the spring I think we will already see several projects with medical connections come to the fore.”
You can read an interview here with SCAPIS Director Göran Bergström about plans for studies to be carried out with data from SCAPIS in AI Innovation of Sweden – LINK.
Mikael Elam, Justin Schneiderman, Lars Lindsköld and Jenny Almkvist also represent the Life Science domain for AI Innovation of Sweden.
Great potential for health care
Artificial intelligence involves using machine learning to train computers to process very large amounts of data. The programs learn from mistakes. Unlike usual statistical computer programs, artificial intelligence uses dynamic algorithms made up of mathematical instructions provided by their programmers.
“The program is not designed to do everything right like a regular computer application, but rather to learn a task through practice. After the software has been well-trained, the machine never becomes tired and can work at inconvenient hours. Doctors and other health care personnel gain time to instead give extra attention to the more complex cases and to meet with patients,” says Jan Kilhamn, head of health care and medical services R&D in Region Västra Götaland.
“We are already seeing AI being widely used for diagnostic imaging – to interpret X-ray and endoscopic images, for example. Eventually we also expect that AI will become part of clinical decision support, improving the precision and speed of diagnostics.”
Several different AI projects are already under way within Region Västra Götaland. One example is the decision support for caries based on X-rays, which involves the Swedish Dental Service. So far the model is accurate 75 percent of the cases, which is on a par with actual dentists. Another example is a project at Södra Älvsborg Hospital from 2017, where the aim was to predict the risk of serious lung inflammation and the resulting extended hospital stay. A model in the project, which investigated how accurately the duration of a hospital stay could be predicted when the patient was admitted, turned out to also work well for other diagnoses, and now a new project is investigating how the model can be used more broadly.
One of the speakers at the opening of AI Innovation of Sweden was Sahlgrenska University Hospital Director Ann-Marie Wennberg. She emphasized the increasing importance of AI in health care and the fact that the initiative is important for both Sahlgrenska University Hospital and others involved in health care.
“AI will be part of our standard treatments and our standard diagnostics,” she said. “We also have something that is important for anyone involved in AI Innovation of Sweden – our patients and also our patient data.”
Ann-Marie Wennberg maintains that AI will simplify and improve a lot in health care: “It will help us with one of the big challenges we have–patient accessibility. We will be able to communication with our patients in a completely different way, and as health care providers we will become more readily available.”
AI Innovation of Sweden’s website: https://www.ai.se/en
SCAPIS website: http://scapis.se/
TEXT: ELIN LINDSTRÖM CLAESSEN / MADELENE SZABÓ
PHOTO: LINDHOLMEN SCIENCE PARK / MALIN ARNESSON/ MADELENE SZABÓ
