Quantifying Gait Changes Using Microsoft Kinect and Sample Entropy

TL;DR

This paper presents a novel method using Microsoft Kinect and sample entropy to quantify gait changes, demonstrating its effectiveness in detecting differences caused by motion-restricting devices and potential for physical therapy monitoring.

Contribution

It introduces a new approach combining Kinect-based joint tracking with sample entropy to quantify gait variability and detect changes in walking patterns.

Findings

Method detects gait differences with and without motion-restricting devices.

Preliminary results show potential for physical therapy progress tracking.

Sample entropy effectively quantifies joint movement variability.

Abstract

This study describes a method to quantify potential gait changes in human subjects. Microsoft Kinect devices were used to provide and track coordinates of fifteen different joints of a subject over time. Three male subjects walk a 10-foot path multiple times with and without motion-restricting devices. Their walking patterns were recorded via two Kinect devices through frontal and sagittal planes. A modified sample entropy (SE) value was computed to quantify the variability of the time series for each joint. The SE values with and without motion-restricting devices were used to compare the changes in each joint. The preliminary results of the experiments show that the proposed quantification method can detect differences in walking patterns with and without motion-restricting devices. The proposed method has the potential to be applied to track personal progress in physical therapy sessions.