Using Artificial Intelligence (AI) to Help Identify Care Partner Burden
For many people living with dementia, living at home with family is preferable to assisted living or long-term care. However, as the symptoms of dementia progress, people living with dementia often experience a decline in functional abilities, which in turn requires greater support from their caregivers.
When informal caregivers – who may be spouses, family, or friends – take on greater responsibilities in assisting people living with dementia complete activities of daily living, they increase their own risk of stress and burnout. Yet it can be difficult for caregivers to self-identify when their load has become a burden, and when they may need to engage additional supports.
Thanks to new funding from the Alzheimer’s Association and Brain Canada, researchers at the Bruyère Research Institute, Carleton University, AGE-WELL SAM3 National Innovation Hub, and Oregon Health and Science University (OHSU) are working to develop a measure, using artificial intelligence (AI), that can help identify when care partners are experiencing higher levels of burden.
To monitor and assess real-time behaviours in a natural environment, the Collaborative Aging Research Using Technology (CART) platform leverages sensors installed throughout an individual’s home. The platform can passively capture activity, such as room transitions, time spent in or out of the house, whether people are in the room together or not, along with health measures, such as sleep patterns.
“We have already been evaluating the CART platform in the homes of people living with dementia and their care partners over the last few years,” said Dr. Neil Thomas, physician with the Bruyère Memory Program and the principal investigator of the project at the Bruyère Research Institute. “Using longitudinal data from larger studies in collaboration with OHSU, we are now looking at developing a digital signature, using AI, to help detect the changes in daily caregiving-related activities that correlate with higher risk of caregiver burden.”
Data generated by home-sensor technologies are representative of health and behavioural changes over time, giving a more accurate picture of someone’s physical and cognitive status, rather than a single clinical evaluation. Additionally, data are less susceptible to recall biases that can occur with self-reporting activities.
“So much is on the shoulders of the care partner,” said Dr. Thomas. “It can be hard to track or recognize gradual changes, especially when you are already overwhelmed. The idea behind this tool is to help collect objective data that can help flag the need for support services before there is a crisis.”
As the Canadian population continues to age, and an increasing number of older adults are diagnosed with some form of dementia, the corresponding challenges for care partners are also becoming more prevalent. Use of home-based sensors can not only help caregivers understand the needs of people living with dementia, but with added AI functionality, data can also help caregivers protect their own health and quality of life.