Torque Analysis of Male-Female Gait and Identification using Machine Learning

摘要

Relevance of gait-attributed changes in male and female subjects could be a significant tool for clinicians to identify and diagnose movement disorders. In this paper, we used 6 low-cost wearable mobile phone sensors to extract gait data. Classification and inverse dynamic analysis were performed to identify gait changes for distinctly identifying gender-specific characteristics. Machine learning algorithms were used to classify the joint kinetic and kinematic parameters. Based on current analysis and in the context wearable low-cost sensors, the change in average torque amplitude and torque differences across right and left hip and ankle could be the relevant classification biomarker.