SWAF: Smart Wheelchair for Fall Risk Assessment

Abstract

Fall prevention is an important healthcare issue for aged societies because a fall accident usually accompanies with multiple fractures and may cause a life-shortening effect. It is especially difficult for elderly people to get fully recovered from a fall accident. Therefore, fall risk assessment for elderly people is highly desirable. A commonly-used assessment method is to measure the time an older person needs to sit down and stand up. The more time the elderly people spend in sitting down and standing up, the higher risk there will be for them to have a fall. In this work, we focus on elderly people who use wheelchair in their daily lives because those people are the most vulnerable group of people who cannot afford to have a fall accident. For this, we developed SWAF, a smart wheelchair for fall risk assessment, which automatically measures the time a user spends in sitting down and standing up. Pressure sensors are deployed at armrests and a surface area of wheelchair pad where a user’s legs and buttocks may touch. To remove noise induced by pressure sensors, a SVM (Support Vector Machine)-based regression filter is adopted. The filtered signals are then used to determine the time interval of sitting down and standing up. To evaluate the accuracy of SWAF, 30 volunteers were invited to carry out a series of experiment. Our results show that SWAF can achieve an average error less than 150ms in measuring the time a user spends in sitting down and standing up.