Enhancing Gait Analysis and Pathway Classification Through Ground Impedance-Based Shoes: An Innovative Approach

Abstract

The type of walking pathway influences gait. Thus, a wearable gait sensing technique is important for continuous gait analysis. However, most of the wearable sensing technologies employed in gait analysis solely provide data on gait parameters and do not have mechanisms to sense and account for the type of pathways. In this article, a novel technique is developed to simultaneously identify some of the spatiotemporal gait parameters and the type of pathway on which the subject is walking. This is achieved by measuring the electrical impedance of the floor between the shoes employing a measurement system reported recently. This article shows that gait parameters can be derived using the impedance values measured between the shoes. These impedance values change as the legs move, primarily due to changes in capacitance between the shoe and the pathway. A suitable algorithm is developed and tested to estimate the gait parameters and walking speed from the developed prototype, and this is compared with the parameters obtained from reference force plate-based sensing. The testing is done on seven human subjects. The average root-mean-square error (RMSE) values for different gait parameters were found to be 0.02 s, 1.4 cm, 2.6 cm, 0.07 s, 0.8 steps/min, 2.24%, and 2.24%, for stride time (STT), stride length (STL), step length (SL), step time (ST), cadence (CD), stance phase (SP), swing phase (SWP), respectively, and a worst-case error of ±5% in walking speed is observed. Further, the human subjects walked on different pedestrian pathways. Different features were extracted from the impedance waveform, which helped in successfully classifying all the six types of pathways we tested.