Sensor project helps people with dementia and their nurses

Early warning system

The shortage of people in the healthcare sector is an increasing problem. You can solve this by recruiting employees from abroad or by looking for technical solutions. Jan Kleine Deters is a professor-researcher at the Electrical Engineering degree programme and the Smart Industry professorship. 'One of the solutions is an early warning system for misunderstood behaviour of people with dementia in nursing homes. Sensors alert nurses if the patient exhibits that behaviour. This is better for the people with dementia and for the nurses who don't have to keep an eye on everyone all the time and therefore experience less stress with their job.'  

The early warning system with sensors is the PhD trajectory of Jan Kleine Deters and part of the MOOD-Sense project. Jan steers the project in the right direction and coordinates fifteen Electrical Engineering students who are doing partial research and five graduate students from Industrial Product Design, Electrical Engineering, and Nursing. The research started in September 2020, and the first pilot is currently running in an institution of the Zonnehuisgroep Noord. The students also come to the nursing home in Loppersum if this is possible with the COVID-19 measures. 

Kleine Deters: 'We try to describe the behaviour of the patients objectively. People who walk restlessly up and down, people who are constantly tapping their watch on the table, people who are constantly calling for the nurse. To map this out, we consult the employees. They have to say which behaviour is problematic. Some walking around the hallway could be out for a nice walk, but it could also be someone who wanders aimlessly for hours and endangers themselves.'  

Smart ring

The researchers use all kinds of sensors that they place in different places. 'Heart rate monitors, light and sound sensors, moisture meters in pillows and mattresses, we also have a smart ring that registers movements. The data coming from those sensors must be interpreted by algorithms, and ultimately the outcome of such an algorithm is that the employee receives a signal on their mobile phone. Then they can take a look.'

Students conduct interviews with employees. They come up with demands for the system that they never could have imagined. 'Employees pointed out to them that smart cushions need to look pretty and need to match with the chairs, otherwise the residents will not use them. The smart rings should closely resemble the jewellery they already wear, possibly in gold or silver. Black or red rings are not that desirable.'

How do you measure anxiety? 

The project faces the necessary challenges. 'How do you measure anxiety? A high heart rate or sweating gives an indication, but they could also be symptoms of something else. Unfortunately, we cannot yet measure the stress hormone cortisol with sensors. We also try to find patterns that allow us to predict when someone is about to cuss. We look for correlations between certain forms of movement and problematic behaviour. But misunderstood behaviour also includes non-physical behaviour. It is a very complex study, and we are only in the pioneering phase At the moment we mainly collect data in the wild in the hope of demonstrating correlations.'  

Demonstrating correlations is an important part of the MOOD-Sense project. 'That is why talking with the employees of the nursing homes is so important. For example, the employees may notice that an agitated atmosphere can arise whenever a certain group of people is sitting together. Or that someone quickly becomes overstimulated by sounds in the environment and then starts cursing, or if someone is understimulated by their environment and starts wandering around in search of stimuli.'

Spurious correlations 

Kleine Deters: 'You have to be very careful with the interpretation of the sensor data and making connections. In a research project in Portugal, sensors registered strange irregular movements in a group of elderly people. That seemed alarming until someone took a closer look and saw that they were playing cards. Other researchers found that a person with dementia was often agitated when the moisture level in their environment was relatively low. Very interesting, they thought at first. But then, after talking to the employees, it turned out that the residents always take a shower before they have a family visit. That family visit was good for the patient's mood, but it had nothing to do with dryness or humidity.'

The project will run for another three years. 'I want to take an important step with system development. I want the knowledge that we gain and the innovations we develop to also end up in the education at Hanze UAS. It is wonderful to work as an engineer on a project that contributes to solutions for a societal issue.' This project is an excellent example of how Hanze Healthy Ageing works within one of its core themes Frailty & Adequate Care.  

Text: Luuk Steemers