Investigation of the Suitability of Utilizing Permutation Entropy to Characterize Gait Dynamics

Permutation entropy (PE) is a natural measure of the complexity of a dynamic system which boasts robustness to noise and computational efficiency. Since its introduction in 2002, PE has served as a valuable tool for assessing nonlinear systems such as speech signals and EEG signals. In this paper, the suitability of utilizing PE as a new method for characterizing walking gait dynamics is explored. Results obtained by analyzing the medial-lateral center of pressure migration of participants during treadmill walking demonstrate that PE shows good preliminary repeatability. PE shows sensitivity to a change in walking conditions; walk alone vs. addition of a concurrent cognitive task. A highly significant age effect in PE (p<0.001) during dual task walking was observed. Analysis revealed no significant age effect on PE for the walk alone condition.