Lichamelijke activiteitsbeoordeling met behulp van draagbare sensoren en machinaal leren

Worldwide, the population is ageing, resulting in an increased share of older adults. It is estimated that between 2015 and 2030, the number of elderly is expected to grow from ~900M to ~1.5B, and is even expected to jump to 2B by 2050. Apart from the increasing number of elderly, life expectancy will also go up because healthcare is improving. As a result, potentially greater numbers of people will be in need of care, thereby challenging healthcare systems and increasing health care expenses. A possible way to reduce healthcare costs in elderly could be by monitoring their physical function. This will inform about physical self-reliance, which is the ability to perform certain activities of daily living, such as personal care, mobility and feeding, independently. The level of physical self-reliance is not only a prominent factor affecting a person's quality of life, but also their use of healthcare services. This suggests that when physical function is declining, and thus self-reliance, healthcare use will increase and so will expenditure. However, when worsening of physical function could be detected timely, individual care planning could be optimized to prevent this. Currently, monitoring of physical function in older adults is not optimal. The main drawbacks of the existing methods are: 1). Being questionnaire-based or relying on (extensive) physical tests. These methods, either do not involve actual physical performance or cannot be performed independently in a home environment; 2). Measurements are only performed intermittently with large intervals. This prevents carers from providing optimized care, by being able to timely adjust care planning. We suggest that the application of wearable sensors and Internet-of-Things devices, which have proven to be promising tools for home-based activity recognition by combining motion and context assessment for example, could overcome these drawbacks. Therefore, the aim of the proposed project is to develop a user-friendly system for independent objective monitoring and evaluation of physical function in community-dwelling healthy older adults. We will focus on validated tools such as the modified Physical Performance Test, which includes a range of basic and complex activities of daily living. If successful, the system may help elderly to stay self-reliant with a high quality of life for a prolonged period, thereby potentially reducing healthcare costs significantly.