Smart Socks for Gait Analysis: A Comparative Study of Algorithms

Abstract

This study aims to evaluate filtering algorithms for extracting spatio-temporal parameters to assess the agreement between the e-textile system Sensoria Smart Socks produced by Sensoria Health and the inertial system Mobility Lab produced by APDM, used as a reference. We collected experimental data from eighteen healthy subjects and conducted a statistical analysis of gait cycle time, cadence, swing, and stance. The signals were conditioned using Butterworth filtering with different cut-off frequencies, following previous research with similar aims and experimental settings. Results show a strong agreement for gait cycle time and cadence, regardless of the filtering method used. However, a higher cutting frequency or no filtering results in a better agreement for the stance and swing, eliminating the significant systematic errors detected with a lower cutting frequency. In the future, it would be desirable to develop hybrid algorithms that can calculate different parameters with different cut-off frequencies.Clinical Relevance - This study establishes the possibility to use an innovative wearable device for rehabilitation purposes.