Healthcare system for non-invasive fall detection in indoor environment

Abstract

Fall incidents and the sustained injuries represent the main causes of accidents for elderly people, and also the third cause of chronic disability. The rapid detection of a fall event can reduce the mortality risk, avoiding also the aggravation of injuries. In this paper an embedded healthcare system based on a microwave radar is presented. A Continuous Wave (CW) Doppler radar is used to detect the changes in speed of different persons experienced during daily activities, namely falling and normal/random movements. The resulted speed signals are then processed in real-time by a digital signal processor (DSP) in order to detect fall incidents. Experimental results, conducted on real human volunteers in a real room setting, have shown a success rate of 100 % in detecting fall events. Moreover, no false positives have been reported.