Gait rehabilitation assessment based on microwave Doppler radars embedded in walkers

Abstract

Human gait is a complex motion that implies the movement of different parts of the body such as arms, legs and feet, being the functional human movements' analysis indispensable for bio-mechanical diagnostic and treatment tool for clinics and rehabilitation services. During the rehabilitation process, walkers are frequently prescribed to improve the patient's stability but can also be transformed into instruments for quantitative evaluation of rehabilitation progress by embedding sensors to capture the motion characteristics. In this work, a practical approach concerning the possibilities to use microwave Doppler radars embedded in four wheels walkers for gait capture is presented. The signals acquired from the sensors are processed using time-frequency transform such as STFT. A set of gait characteristics, such as gait velocity and stride rate, are extracted based on wavelet signal processing, STFT spectrogram and moving average filtering. A set of spectrogram features is calculated to discriminate between normal and abnormal gait.