Validation of a Passive Dynamic Walker Model for Human Gait Analysis

Ismet Handžić, University of South Florida
Kyle B. Reed, University of South Florida

Abstract

While the dynamics and mathematics of passive dynamic walking (PDW) models have been extensively researched, it has not been until recently that they have been used for practical applications in rehabilitation and gait analysis. In this study, we evaluate the validity of using a two dimensional PDW for human gait analysis. Here, a PDW model is compared to recorded kinematic and kinetic walking data under normal conditions. We also study asymmetric gait by imposing a shank mass asymmetry and comparing it to a PDW model under the same asymmetric conditions. Kinematic and kinetic data for normal gait was taken from one subject using a motion capture system and a force plate, respectively. Gait under the asymmetric shank mass conditions was recorded by measuring the drifting radius of curvature from five participants walking blindfolded with a mass attached to their shank. While the PDW model lacks ankles (dorsiflexion), joint damping, and joint stiffness, the kinematics, kinetics, and gait asymmetry were comparable. Kinetic comparisons show agreement in general ground reaction force magnitude and profile. Kinematic results yield a good match in temporal and spatial gait characteristics. Asymmetric analysis of gait demonstrated that the PDW model can accurately predict the direction of step asymmetry.