Presentation Title

Generating Real-Time Musculoskeletal Models Using a Motion Capture System

Start Date

November 2016

End Date

November 2016

Location

HUB 302-#57

Type of Presentation

Poster

Abstract

Musculoskeletal modeling has been widely used to study human movement and to aid in treating musculoskeletal diseases. Musculoskeletal models can be combined with human motion capture data to analyze movement and generate subject-specific dynamic simulations. Currently, there is a lack of integration of real-time musculoskeletal simulations using motion capture data. This limits potential applications such as motion training and analysis of movement disabilities. In this project, we aim to develop a system using a computational framework (OpenSim) and motion capture (OptiTrack) that generates musculoskeletal models in real-time. Three subjects had 19 reflective markers placed on their body in various locations, and were instructed to walk on a treadmill at their preferred walking speed. A 6-camera marker-based OptiTrack system was used to capture their motion and live-stream that trajectory to a MATLAB program. MATLAB used a 4th order Butterworth filter with a 15Hz cut-off frequency to filter the data, and applied inverse kinematics, inverse dynamics, and forward dynamics operations to that data. We expect the filtered data in MATLAB to then be streamed to the OpenSim software to generate a subject-specific simulation. We also expect this system to generate a real-time model from OpenSim using the OptiTrack program and the MATLAB interface. This system can be used to advance the generation of real-time simulations of various models while reducing the time required to process and analyze the results. The real-time models can also be used to improve training for athletes and provide biofeedback for patients with musculoskeletal disorders.

This document is currently not available here.

Share

COinS
 
Nov 12th, 4:00 PM Nov 12th, 5:00 PM

Generating Real-Time Musculoskeletal Models Using a Motion Capture System

HUB 302-#57

Musculoskeletal modeling has been widely used to study human movement and to aid in treating musculoskeletal diseases. Musculoskeletal models can be combined with human motion capture data to analyze movement and generate subject-specific dynamic simulations. Currently, there is a lack of integration of real-time musculoskeletal simulations using motion capture data. This limits potential applications such as motion training and analysis of movement disabilities. In this project, we aim to develop a system using a computational framework (OpenSim) and motion capture (OptiTrack) that generates musculoskeletal models in real-time. Three subjects had 19 reflective markers placed on their body in various locations, and were instructed to walk on a treadmill at their preferred walking speed. A 6-camera marker-based OptiTrack system was used to capture their motion and live-stream that trajectory to a MATLAB program. MATLAB used a 4th order Butterworth filter with a 15Hz cut-off frequency to filter the data, and applied inverse kinematics, inverse dynamics, and forward dynamics operations to that data. We expect the filtered data in MATLAB to then be streamed to the OpenSim software to generate a subject-specific simulation. We also expect this system to generate a real-time model from OpenSim using the OptiTrack program and the MATLAB interface. This system can be used to advance the generation of real-time simulations of various models while reducing the time required to process and analyze the results. The real-time models can also be used to improve training for athletes and provide biofeedback for patients with musculoskeletal disorders.