He is passionate about technology. And AI. “Never have I ever experienced anything as mind-blowing,” says Hans Forsberg, about today’s possibilities with neural networks and AI when it comes to areas such as healthtech.
TEXT: ANNIKA ÅKERVALL
During the Inn2Health workshop Cross Collaboration Day, we got to experience an exciting review of the basic factors in the art of building a smart machine, using AI.
“Neural networks are a new way to harness the power of computers. To get the computer to do what you want it to do. You could say that it’s like filling a large box with millions of criteria that must be fulfilled,” says Hans Forsberg, AI Architect at ÅF.
“If you are going to write different categories, for example, if this is fulfilled and this is fulfilled and so on, then we can get the computer to accept that it is so. Given that we have examples to show.”
A lot of data is important
“The training process can take time. But it goes fast if you have new tricks to solve problems,” says Hans Forsberg.
He gives several examples of application areas. Like when an elderly person wants to continue living in her/his home. An application could then be to monitor the number of people and the movement in the home, for example, via cameras and AI technology. This could increase the feeling of security by having relatives follow this. With an agile development, you can start on a small scale and develop gradually.
“It is a paradigm shift – a whole new way to tackle problems. Around the AI concept, there are many fantastic mathematical tools that can be used advantageously,” says Hans Forsberg.
Through his experience of error reporting, he knows that it is important HOW you ask the question. By analysing texts that customers have submitted in a bug report, the computer could find links between what was written, and which box should have been checked, based on the type of case. And connect this to care visits.
“Based on very anonymised ‘training data’ from previous patients, it was possible to build an algorithm to steer the message to the right department. So that we automatically get an indication of what the issue is or how critical it seems,” says Hans Forsberg. Or when this type of patient tends to miss pre-booked appointments. This can inspire the discussion around solutions to this costly dilemma.
The development of artificial intelligence creates opportunities throughout the entire value chain in life science. There is a large variation in the development of digital services in West Sweden, although solutions for decision support strongly dominate among the companies, according to Sahlgrenska Science Park’s survey on Artificial intelligence in life science 2018.
Welcome to download our AI report: Artificiell Intelligens inom life science (in Swedish)