Fall detection and risk of falling assessment with wearable sensors

Abstract

We demonstrate a wearable sensor system for automatic detection of falls and assessment of risk of falling for the elderly through continuous physical activity monitoring. The demonstrated approach uses data measured by a wearable sensor, PAMSys™, for long-term physical activity monitoring. 3-dimensional acceleration data are analyzed to detect falls of a person in order to inform caregivers of such events. Furthermore, as a preventive mechanism, we propose to assess a person's risk of falling using physical activity information. Relevant physical activities include postural transitions, gait initiation, turning, and history of falls. This approach can enable early detection and identification of patterns indicative of high risk of falling in at-risk elders and allow development of more effective preventive measures. The proposed system has the potential to enhance the quality of life and reduce the overall cost of care for elderly persons by assisting them to maintain an independent living style.