MusculoSkeletal Modeling Using Kinect Data For Telerehabilitation

TL;DR

This study explores using affordable Kinect sensors with OpenSim to analyze musculoskeletal parameters for telerehabilitation, focusing on balance assessment and personalized therapy in diverse populations.

Contribution

It demonstrates the feasibility of using low-cost Kinect data with OpenSim for joint kinematics and identifies key biomarkers for balance in telerehabilitation.

Findings

Kinect data can be effectively used with OpenSim for musculoskeletal analysis.

Pelvis list and tilt moments are key biomarkers for balance.

Significant differences in joint forces and torques across BMI categories.

Abstract

Balance, gait and postural control are some of the key factors in determining the overall stability of an individual. Several highend and costly solutions exist to perform movement analysis in clinical settings. OpenSim is a tool which uses 39 marker positions, obtained from such highend solutions like VICON or equivalent multicamera setup, for the analysis of inverse kinematics and inverse dynamics. However, an affordable solution for deriving musculoskeletal joint kinematics parameters using a low cost Kinect device is of immense importance. In this paper, we initially study the feasibility of using OpenSim tool on 20 joint locations of human being, obtained from Kinect data. Next, we analyze the various joint forces and torques experienced during a Single Limb Stance (SLS) exercise performed by healthy subjects in normal, overweight and obese categories. Results indicate that a subset of parameters related to forces and torque in hip, lumber and pelvis are the most important ones that contribute significantly in maintaining static balance in SLS. Statistical analysis demonstrates that the pelvis list and tilt moments are the key biomarkers for maintaining the statibility in SLS, thus leading to a possibility of personalizing the therapy in tele-rehabilitation.