Sleep problems can have a major impact on cognitive function, particularly in older adults who are also more likely to have a clinically diagnosed cognitive impairment. While several approved methods for sleep disturbance exist, most are not suitable for use within the abovementioned population. Often the reason for this is due to the symptoms associated with cognitive impairment or for the invasiveness of some sleep profiling methods. Developed from current sleep actigraphy techniques, this paper presents a non-contact alternative method for sleep profiling that is deemed to be more suitable for long term monitoring in the older population than current clinically approved techniques. This first evaluation has been conducted with a young control group in order to validate the approach. The results have demonstrated that based on the approach a random forest classifier using features calculated from optimally placed static accelerometers can produce a sleep/wake classification accuracy of 92%.
|Title of host publication||Impact Analysis of Solutions for Chronic Disease Prevention and Management, Lecture Notes in Computer Science|
|Place of Publication||Artimino, Italy|
|Publication status||Published (in print/issue) - 2012|