"Collecting data has become very valuable for a lot of different companies. If you analyse them well, you can discover things you normally wouldn't know."
Data sets, machine learning and algorithms. What do they have to do with healthcare? At UT, everything. Here, you collaborate on innovations in which tech, behaviour and health come together. Bert-Jan van Beijnum, Associate Professor in Biomedical Signals and Systems at UT, and Computer Science student Jay Tabudlo both agree on the impact data can have on society. For example, with an app for diabetes type 2 patients.
Jay, you decided to continue your studies with a master’s degree at UT, why?
"My education in Software Engineering at a university of applied science – ‘hbo’ in Dutch – was very interesting, but also pretty general. I wanted to dive deeper into big data, or machine learning. And that’s what I got in UT’s Master’s in Computer Science. During my courses, I get to work on different types of data sets. It’s so cool to see how it all works, and to become more aware of all the upcoming developments and possibilities in this field. If you don’t want to start working right away and you like to challenge yourself, a master’s like this is a very good choice. And, of course, Computer Science comes with a job guarantee, while also being broad enough to prepare you for whichever direction you want."
In fact, Associate Professor in Biomedical Signals and Systems, Bert-Jan van Beijnum, encourages students to pursue their personal interests. "At UT, every student is at the helm of their own assignments. We encourage them to follow their own ideas. Being able to discover more on the topics that match your personal interests makes studying interesting and challenging."
How do you see the changing role of data?
Jay: "Nowadays, we are collecting far more data than we did ten years ago. It has become very valuable for a lot of different companies. You can achieve a lot with data: if you analyse them well, you can discover things you normally wouldn't know. It is not always used for the right purposes, but they can certainly be put to very good use."
One noble cause is a project Bert-Jan is currently working on: the Diameter app. It is an eHealth coaching system for patients with diabetes type 2. The app monitors users’ lifestyle, exercise, food diary and glucose levels. Bert-Jan has been involved since the start, and he’s passionate about the possibilities. "With the right sensors, calculations and analyses, this app can offer real support to health professionals."
How can machine learning influence patients positively in their daily lives?
Jay: "Manually looking through data like the ones involved in the Diameter app takes a lot of time and effort. An algorithm can help – and is often much more accurate. For example, if a blood value is too high, the right algorithm in the app can find that out and send a message to the user or the healthcare professional."
Bert-Jan adds: "If we read the signals from sensors properly and use accurate models, we can give patients specific advice to get their blood sugar at the right level, or to keep it there. But before we reach that stage, we need to know which parameters are influencing factors. Eating behaviour is obviously one example. But think about physical activity or sleep patterns, too. Eventually, I hope we can give every patient personalised advice based on individual preferences, habits, and physiology. Before we reach that goal, we have a lot of analysing, model making, micro-interventions and machine learning to do."
Are tech and health not two very different areas of expertise?
Bert-Jan: "I didn't know much about diabetes when I joined this research. In such projects, it is the challenge to learn and understand each other, so that different worlds can connect. As a technician, I had to discover the world of an internist, and to form a picture of a nurse’s job who has daily contact with patients. Conversely, a health professional has to get a feel for what is needed to make a reliable model. It takes time and dialogue, but together we can do beautiful things. UT is very strong in connecting technical-scientific researchers and clinicians. Our international colleagues in other countries are very surprised when they hear that I, as an electrical engineer, am in touch with doctors and patients. Most data in scientific research is collected clinically; we measure patients during their daily lives. That’s why we’re about to start a trial with 800 patients from the ZGT Hospital Group Twente."
Jay recognises the added value of combining strengths. "At UT, you can follow any master’s course you want, so several courses are with students from different study programmes. As future Computer Scientists, we’re good at programming, but it sure helps if other students bring, say, their mathematical skills to the table!"
What possible impact can apps like the Diameter have?
Jay: "I wouldn't be surprised if we start working with apps like the Diameter more and more in the future. I think they will become very important. That's why I like this master's programme so much; I'm learning things with which I will really be able to contribute to society later."
Bert-Jan agrees: "The social relevance makes these types of projects so interesting. Of course, diabetes is a problem, but if patients could reverse the process of their illness with an intervention like the Diameter app, the impact on society would be huge. For me – and for most students – work is so much more fun and rewarding when it becomes concrete like this. In our area of expertise, you can be dealing with real societal challenges."