BioMechanics Lab

Edit SideBar
Homepage Maintenance

Research

Introduction

We focus on three topics - (1) Musculoskeletal simulation, (2) Joint kinematics, and (3) Skeletal feature identification

The following contents are being edited.

1. Human Musculoskeletal Dynamics Simulation

The knee receives about three times body weight during walking and seven times body weight during stair descending. This is because the muscles attached to bones pull them together and increase the internal force. We can move our body by contracting muscles which are the only actuators in our body. From the computational mechanics point of view human body is a multibody system composed of multiple rigid bodies (bones or segments) and actuators (muscles). Connective soft tissues such as ligament and joint capsule acts as passive elements, and constrain the range of motion of a joint.

There have been many studies on mathematical modeling of muscle and ligament such as force-length and force-velocity properties. All combined we can build a human musculoskeletal multibody system - bones, joints, and muscles. This is a multibody system like a car or robot. We can use the well established theories in multibody system dynamics to understand forces and moments in joints and tensions in muscles during human activities. Commercial and non-commercial solvers exist for general multibody system dynamics along with specific solvers for human musculoskeletal system dynamics.

Research in our laboratory focuses on joint injury mechanisms such as the degenerative diseases in the knee, ankle and spine by mechanical forces. While understanding the joint internal force is necessary to understand the pathomechanics of joint injuries calculating (or predicting) joint internal force is not as easy as you think. We still do not well understand the way human control their muscles during activities, which is called neuromuscular control. If you are controlling a robot energy efficiency or certain performance metric should be defined to find the optimal control. Human research is very exciting because there are still many things to learn and to do.

Link to: Estimation of knee internal force
Link to: Estimation of ground reaction force
Link to: Development of human thoracic musculoskeletal model
Link to: Estimation of ankle internal force and moment

2. Motion Capture of Human Skeletal Kinematics

Kinematics should be obtained for kinetics analysis. Body position, velocity and acceleration should be obtained first to calculate forces and moments on the body. Accuracy of kinematics influences the results of kinetics. Motion capture system has a long history in arts and movies. Modern motion capture system called optoelectronic motion capture system hires reflective ball markers, infrared light emitters and CCD cameras with infrared filters. This system can achieve high accuracy (< 1mm) and high speed (>100FPS) in tracking the ball markers.

In quantifying human body kinematics the markers are placed on the skin of body segments thus the markers undergo skin vibration and deformation, which is called the skin tissue artifact in motion capture and degrades the accuracy of the kinematics. The relative movement between an external skin marker and its underlying bone can be in the order of 10 to 20 mm during walking. Many different cluster methods have been developed to obtain rigid body motions of body segments to filter out the skin noise by putting multiple markers for each segment.

X-ray imaging has a long history since Roentgen discovered the existence of X-rays in 1895. Modern digital imaging techniques enabled the digital recording of continuous X-ray images, called the fluoroscopy. The portable C-arm is a popular embodiment of the fluoroscopic system. Single C-arm was previously used to calculate the three-dimensional position and rotation of a bone or prosthesis. Dual C-arm, or bi-plane fluoroscopic system, has been applied to joint biomechanics to quantify the human skeletal kinematics and track total joint protheses. This state-of-the-art skeletal motion capture system is increasing its application to all important joints in human body and development of artificial joints.

Link to: Bi-plane fluoroscopic system at Chung-Ang University
Link to: Combination of bi-plane fluoroscopic system and optoelectronic motion capture system
Link to: Validation of cluster methods using our motion capture systems
Link to: Knee skeletal kinematics
Link to: Ankle skeletal kinematics
Link to: Cervical skeletal kinematics

3. Human Skeletal Geometry Analysis

The rigid body kinematics of body segments give the complete configuration and movements of body segments. Quantification of rigid body movements of the human body requires anatomical coordinate systems in each body segment. There have been some agreements in determining the anatomical coordinate systems for bones in important joints such as the knee, spine, shoulder and ankle. This anatomical coordinate system is based on anatomical features on the bone, whose reproducibility is influenced by the observer.

The shape of the bone is not easy to define. The history of human anatomy has put a great effort into determining names of distinct shapes on bones. Nevertheless, there is ambiguity in determining positions on the bone. Geometric features such as eminence and curvature are frequently used but these are not as clear as you think and have low inter-observer reproducibility. This reproducibility may degrade according to observer's experience.

Statistical shape modeling is a technique to use a database of polygon bone models to calculate a mean shape and eigen shape vectors. Thus a target bone shape can be approximated as a combination of the mean shape and the eigen shape vectors. This has shown good results in modeling human bone and automatically determining anatomical landmarks on the bone.

The predictability of statistical shape modeling increases with the size of its database. It is somewhat laborious to prepare for a large database of bone polygon models. We are developing a template-based shape matching method. This is based on calculation of shape correspondence between a template bone and target bone. This method is applicable to any type of bone as long as we have a template bone. This automatic feature detection of bone has a great applicability in human anatomy, biomechanics and robotic surgery.

Link to: Statistical shape modeling of the femur to determine anatomical landmarks
Link to: Statistical shape modeling to determine anatomical features
Link to: Template-based shape matching and determination of anatomical landmarks

http://cau.ac.kr/~skoo/cgi-bin/pmwiki/pmwiki.php?n=Work.ResearchOld20140905
Lab. Reference

Edit
Page last modified on October 05, 2016, at 09:05 AM