Smart wearable device for real time gait event detection during running

Abstract

Gait studies in sports and rehabilitation may benefit from online gait event detection algorithms for use in event-dependant feedback strategies. Event-dependant feedback systems may further benefit from durable, lightweight, low cost sensors for gait event detection. In this regard, this study describes the development and feasibility evaluation of an online gait event detection system using inertial sensor technology for the identification of Heel Strike (HS) and Toe Off (TO) events during treadmill running. Custom developed system software performs the online data acquisition, processing, graphical representations of lower extremity kinematics and online gait event detection. For increased robustness, a Finite State Controller architecture is employed for continuous detections of HS and TO during running. Pilot tests conducted with 7 healthy subjects during treadmill running verified the accuracy of gait event detection with mean timing errors of 14ms for HS and 27ms for TO compared to normative values. The effectiveness and robustness of gait event detection is promising signifying the use of the system for triggering event-dependant feedback during running gait retraining.