Artificial intelligence developed at the University of Agder is now being used to review 29,000 scientific papers on COVID-19. The goal is to create a knowledge base for health professionals.
“We are pleased that our technology can help prevent the spread of the coronavirus. Our technology can process large amounts of data and quickly create an overview of a great deal of information”, says Ole-Christoffer Granmo, professor of artificial intelligence at the Department of Information and Communication Technology at the University of Agder (UiA).
Granmo is director of the Centre for Artificial Intelligence Research (CAIR) and founder and owner of the company Anzyz Technologies in Grimstad. Anzyz Technologies owns the technology that is now being used and has been commissioned to develop the service.
The technology of Anzyz Technologies can teach a computer to understand different languages and quickly read through 29,000 scholarly articles on coronavirus.
“In a matter of days, our machine will have insight into and overview of a specialist topic - in this case, the coronavirus”, Granmo says.
The aim is to provide health professionals with a knowledge base where they can find an overview of symptoms and map the risk different patient groups have of being infected. There is a great need for help to determine if someone is infected with coronavirus or not.
Granmo explains that it is challenging for healthcare professionals to get an overview of all the symptoms and risks of COVID-19. A lot of research on this is being done, and so far there are 29,000 articles about the virus.
Anzyz Technologies has previously collaborated on research with the Hospital of Southern Norway (SSHF), where they together developed a clinical diagnosis system for doctors. The technology was used to find out if a patient had a drug allergy in connection with surgery.
“Now we use the same technology to read through the articles on COVID-19. In the project with SSHF, the artificial intelligence technology and our machine learned the medical terminology. The technology still remembers this, and that makes the work faster this time”, says Granmo.
In collaboration with hospitals in Norway and Microsoft, the first prototype is now complete.
“Now it comes down to making it as good as possible in the most efficient way. Fortunately, the response and the cooperation of the hospitals has been very good”, he says.