“A 60% risk of falling is only useful if care providers know how to respond to it.”
Background
Nursing home care in the Netherlands is under increasing pressure. As the population ages, the demand for care is growing, whilst the number of available care professionals is declining. At the same time, care is becoming more complex. Innovative technology can play a key role in better supporting care professionals and improving the quality of care for residents.
This project was initiated by Momo Medical, which has developed a bed sensor for elderly care facilities. Until this project, the data from these sensors had remained largely unused; in this project, it is being converted into usable data for practical applications.
What are predictive models in healthcare?
In this project, smart models are being developed that predict whether someone, for example, is at increased risk of falling, developing pressure ulcers or requiring more care. The models use data from bed sensors placed under mattresses. These sensors measure, among other things, movement, heart rate and sleep patterns.
By combining this data with existing healthcare data, a comprehensive picture of a person’s state of health emerges. Based on this, patterns can be identified that indicate increased risks. For example: someone who sleeps less well and gets up more often may have a higher risk of falling.
The project focuses not only on making these predictions, but also on how healthcare professionals can use this information. When should you intervene? What does a particular risk percentage mean? And how do you ensure that technology continues to support good care?
The results are particularly relevant to elderly care, where prevention and early detection can make a significant difference to quality of life and help reduce the workload.
Partners in this project
- Tante Louise
- Fontys
- Tilburg University
- Avoord
- Qcare
- Kinetic analysis
- Eindhoven University of Technology
- Momo Medical
- MindLabs
Approach and methodology
The project utilises a combination of:
- Data analysis and AI model development based on sensor data and healthcare data
- Collaboration with healthcare institutions to access real-world data
- Integration of various data sources (bed sensors, reports, medical records)
The greatest challenge lies not only in the technology, but above all in:
- the secure and legally compliant sharing of data
- building trust between partners
- translating insights into practical applications
Intended outcomes and relevance
The project aims to gain an understanding of predictive factors for health risks and methods for effectively combining different data sources. This will lead to the development of AI predictive models and the enhancement of existing healthcare software with predictive capabilities. Ultimately, this should result in fewer falls and pressure ulcers, better and more timely care interventions, and a more efficient use of healthcare capacity.
Contact
Kirsten Geurtz?