Integrating Wearable and personal data using machine-learning approaches
Approved Research ID: 63790
Approval date: October 26th 2020
1. To understand what the normal values of: physical activity through wrist worn trackers and muscle loss through scans are in a large sample of normal population.
2. To understand the relationship between the underlying factors that leads to an individuals' physical activity and muscle mass from their background and lifestyle (height/weight/diet).
We know that many patients, as they become less well, move less and change their pattern of walking. This is seen in the population of brain cancer patients whom we care for. Weakness due to cancer interfering with normal brain function, or from the stress of the cancer on the body (leading to muscle loss) impacts not only quality of life but increases risk of treatment complications.
Current methods of judging a person's functional ability are crude, requiring patients to provide a limited account of what they have been able to accomplish recently. This can be hampered by cultural factors such as language barrier, as well as medical ones, such as poor memory after brain radiotherapy. Not obtaining a clear picture of an individual's functional ability can lead to the wrong medical decisions being made.
We want to transcend this by better understanding how impairments can affect a patient, and in order to do so we need better understand what normal is. Technology has advanced to the point where accelerometers which are present in phones or fit-bits can track movement reliably, and routine scans cancer care provides information on more than the cancer itself. By understand normal variation of physical measures and how they may be affected by an individual's ability before they become sick, then we can use them to pre-empt treatment or support before a patient starts to notice a deterioration.
Public Health Impact:
Our research will be published in peer-reviewed journals, which means that the standards and references we find are relevant will be applied to clinical work across a broad range of subjects. Within our group, we hope to use the data as part of trials to detect advancement of cancer when scans are often misleading. In other fields, such as geriatric medicine, the loss of muscle due to age is a big healthcare burden and having better understanding of how scans and movement link together can form the basis of a more individualised approach in how to address this.