Assessment of a Wireless Accelerometer System for Tracking Asymmetries in Gait

Abstract

This paper presents an approach to tracking asymmetries in gait using a lightweight, inexpensive wireless accelerometer. A review of the current status of gait tracking using inertial sensors shows that wearable sensor systems have potential applications in rehabilitation therapy and sport performance to correct compensatory gait patterns after injury or to prevent overuse injury resulting from asymmetries. The proposed design uses a 9-DOF accelerometer/gyroscope/magnetometer that is capable of logging acceleration data at a user-specified sampling frequency. Raw data is transferred from the sensor system through Bluetooth Low Energy (BLE) wireless communication to a mobile device running an iOS application. Further data processing, filtering, and analysis is conducted in Matlab. The performance of the proposed system is evaluated through a noise analysis and the tracking of a physical pendulum. A low-pass Butterworth filter is designed and implemented to reduce high frequency noise present in raw acceleration data. Finally, a gait-tracking experiment is conducted by strapping a wireless sensor to each foot. Asymmetry is artificially induced through a half-shod condition. Preliminary results indicate that summary statistics from total acceleration measurementsspecifically standard deviationare enough to detect asymmetries in stride.