A Robust Method for Detecting Steps in the Frail Elderly

Physical activity (PA) is strongly related to a person’s quality of life and health, and hence there is a great interest in activity monitoring. Accelerometers are being increasingly used in studies of PA among older adults, however when used on elderly with an abnormal gait the interpretation of the data becomes difficult. The gait of frail older adults and individuals with neurological disorders can be characterized by being slow, shuffling and cautious. As a consequence the signal amplitude of the acceleration data and the gait pattern will differ from healthy individuals. From earlier studies examining the validity of commercially available accelerometer-based body worn sensors it is revealed that the software that follows the system exhibits inadequate validity in step recognition on short walking distances and at slow walking pace. The aim of the study was to enhance the sensitivity of the system to detect steps in the frail elderly.