ENGINEERING TOOLS & MOVEMENT ANALYSIS COMBINE IN BALANCE & POSTURAL CONTROL STUDIES

by Elizabeth Hsiao-Wecksler

Assistant Professor, Human Dynamics and Controls Laboratory,

Department of Mechanical and Industrial Engineering, University of Illinois, USA

The Human Dynamics and Controls Laboratory (HDCL) in the Department of Mechanical and Industrial Engineering at the University of Illinois at Urbana-Champaign, USA was established in July 2002. It is under the direction of Assistant Professor Elizabeth Hsiao-Wecksler. The purpose is to apply engineering tools (dynamic systems modeling, and control theory) and movement analysis to explore issues related to musculoskeletal biomechanics. Dynamics and control models based on inverted pendulums and random walk theory are used to investigate underlying neuromuscular and control mechanisms. The lab's main focus is to understand how balance and postural control change with age, exercise, and condition (e.g., pregnancy). Current projects are examining how balance and postural control are affected during tasks such as quiet-stance, normal and obstructed gait, and unexpected perturbations such as tugs at the waist. Our work is related to learning how falls in older adults can be prevented by studying age-related changes in balance and response to unexpected perturbations. We are also interested in exploring how Tai Chi practice might influence balance and movement strategies. Tai Chi has been found to reduce the likelihood of repeat falls in older adults. Another project is examining the biomechanics of aggressive in-line skating, determining how this exercise might use eccentric contraction of the lower extremity muscles to decelerate the body and reduce impact forces, while maintaining balance on a narrow surface. We are also investigating how balance changes throughout the nine months of pregnancy and up to six months post-partum. The newest project will be examining the biomechanics and torque-generating capacity of the wrist while playing foosball.

The HDCL has a 6-camera Vicon 460 motion analysis system and a 16-channel Delsys EMG system. An oversized (60cm x 90cm) AMTI force platform is embedded into a 32 ft elevated walkway. The Vicon system is the main data collection system, using the motion capture data for kinematic analysis and inverse dynamics analyses with reaction force data to compute joint torques. The workstation is also being used to collect auxiliary analog signals, i.e., force plate, load cell, and EMG data, which may or may not include motion capture. The lab also features assorted load cells and a recent addition in the form of a digital video capture capability which has been used in the skating project. I have found the equipment useful for class projects in a course that I teach - "Modeling of Musculoskeletal Biomechanics". Advanced undergraduate and graduate students use kinematic analysis and inverse dynamics to perform simple analyses and create simple mathematical models (e.g., using springs and dashpots) to represent different human movements, like jumping and walking.

Our plans for the future are to further develop each of our main projects. We hope to demonstrate that the impulse input/ impulse response paradigm will be a useful tool for assessing balance and postural control during a perturbation for a variety of test conditions and test groups. We will continue to investigate the benefits that Tai Chi can provide to both old and young adults. We hope to provide more insight into balance issues experienced during pregnancy, possibly gaining more information that would reduce the likelihood of falls or loss of balance during pregnancy. It would be great if we could translate our findings, on how skaters are able to maintain balance or prevent injury in the event of a fall, to fall and injury prevention of older adults.

A Vicon screen shot showing a stick figure and digital video image of a skater sliding ("grinding") on the grind rail. This skater is also attached to EMG.

The following project details are typical of current work in the lab:

Project 1:
Investigators: E.T. Hsiao-Wecksler; graduate students Arun Ramachandran and Kelly McHugh; collaborator Karl S. Rosengren (Departments of Kinesiology and Psychology).
Title: Effect of Tai Chi on postural control and response strategies.
Description: Tai Chi has been promoted to older adults as an exercise to improve physical and mental fitness. It has also been found to reduce the likelihood of falling in senior citizens. This project explores how Tai Chi experience may modify postural control mechanisms and movement strategies during stance and unexpected external perturbations to balance. Dynamic systems modeling, control theory, and movement analysis are used to examine these issues. Two studies are currently in progress. One is a cross-sectional study, with healthy adults aged 25-63 years, exploring the effect of long-term Tai Chi experience (2-15 years) on standing balance, and gait and obstacle crossing behaviors. The other is a longitudinal study with older adults (65+ years) that is examining how balance and movement strategies may change as a result of five months of Tai Chi training. We are collecting, from the AMTI force plate, ground reaction force data during the gait trials and center of pressure data during the standing balance trials. Full-body motion data from the Vicon system is used to compute lower extremity joint angles and foot positions while crossing the obstacle. Data collection for the cross-sectional study is complete. We are currently analyzing the data. The longitudinal study is in progress.

Project 2:
Investigators: E.T. Hsiao-Wecksler; graduate student Arun Ramachandran.
Title: Postural control during mild impulsive perturbations.
Description: Investigating how individuals respond to disturbances to balance is essential to improving our understanding of the etiology of falls. Balance and postural control mechanisms during perturbed stance may change with age. These differences may manifest themselves in the behavioral characteristics of the postural response noted immediately after a perturbation. We are particularly interested in the response of the postural control system after a transient perturbation. Limited work has been done to explore postural responses to sudden, impulse-like perturbations. In this investigation, the impulse loading and impulse response control-theory paradigm will be used to examine the postural response to a mild, quick-release backward tug. While impulse response and its associated characteristics are rudimentary concepts in engineering control theory, we have only just begun to extend this paradigm to investigate postural control. The purpose of this study is to learn more about how to characterize responses to a transient perturbation, what these responses tell us about the postural control system in general, and how these responses may vary with age. We will be collecting center of pressure data from the AMTI force plate, lower extremity muscle activity data from the Delsys EMG system, and full-body motion data from the Vicon system to compute the center of gravity of the body. This project is currently in the development stage.

Project 3:
Investigators: E.T. Hsiao-Wecksler; undergraduates Molly Hathaway, Susan Shah, Alisa Riskus.
Title: Variations in balance and postural control throughout pregnancy and up to six months postpartum.
Description: Pregnant women anecdotally state that balance changes as pregnancy progresses and the circumference of the trunk and body weight increase. However, no studies have examined how balance, and postural control that moderates balance, may vary throughout pregnancy and the subsequent postpartum period. This study will assess how balance and postural control may vary as a consequence of pregnancy by examining how a subject's postural sway varies over the nine month pregnancy and a following six month postpartum period. This study only uses center of pressure data collected from the AMTI force plate via the Vicon workstation. This study is currently in progress. Fifteen pregnant women and fifteen non-pregnant female control subjects are being tested.

Project 4:
Investigators: E.T. Hsiao-Wecksler; graduate student Matt Major; collaborators Armand Beaudoin and Peter Kurath (Dept of Mech & Industrial Engineering).
Title: Biomechanical analysis of aggressive in-line skating: landing and balance on grind rail.
Description: Aggressive in-line skating is a sport that emphasizes balance. A popular activity is grinding, where the skater jumps onto a grind rail - which may be a specially designed structure at a skate park, or a common handrail on a staircase. In grinding, skaters jump up and accurately place their skates on the rail, smoothly decelerate, and balance upon the rail while sliding (or "riding it out"). In-line skaters have developed a heuristic approach to training. Inherent to their training are exercises that emphasize the development of muscle control during eccentric, muscle stretching contractions to smoothly decelerate the body. For example, before performing a grind, the skater would repeatedly jump upon an object and "stall" - that is, jump, place skates on the rail, decelerate, and hold that position. Our main focus is in the prevention of complete loss of balance, falls, and injury in the event of impact with the ground. In this novel study, we are collecting data on limb motion and forces developed during deceleration activities, such as grinding and stalling. By performing controlled jumping and balancing experiments, this project allows us to gain insight into how these individuals are able to use eccentric contraction to assist with maintaining balance and, perhaps, minimizing impact force and energy. In the lab we have fabricated a steel grind rail (5 cm diameter pipe, 178 cm length x 26 cm height), which is constructed to behave like a simply-supported beam (i.e., sustain no moments at the supporting ends). Load cells at each end of the rail record applied vertical reaction forces. These signals are collected by the Vicon workstation. A full-body marker set allows us to compute the center of gravity of the entire body and joint angles. Using inverse dynamics, we can compute the joint torques in the lower extremities during grinds and stalls. We currently have tested six male skaters (aged 18 - 46 years, 1.5 - 10 years experience). We will be adding EMG data in future testing.

The following is a conference abstract, to be presented to the International Society of Electromyography and Kinesiology (June 18-21, 2004 in Boston, MA), that discusses the results from a stall experiment:
Introduction: Aggressive in-line skating, or trick skating, is a relatively new sport, with only epidemiological data of skate park injuries (Everett, 2002). This novel biomechanical study of aggressive skating focused on the "stall", a balance training activity where the skater repeatedly jumps on an elevated rail and tries to maintain balance. We hypothesized that, to smoothly decelerate the body and minimize impact forces, this exercise emphasizes development of muscle control through eccentric contractions of the lower extremity joints. We also hypothesized that more experienced skaters would minimize impact force.
Methods: Six male skaters (aged 18 - 46 years, 1.5 - 10 years experience) performed 10 stalls each. Each subject was instructed to jump onto a grind rail, maintain balance (approx. 1-3 s), and jump down. No constraints were placed on landing and balancing techniques, allowing the skater to maintain his unique style of performance. A simply-supported, steel grind rail (5 cm diameter pipe, 178 cm length x 26 cm height) was constructed to sustain no moments at the supporting ends. Peak vertical impact force (as % body weight) was recorded and sampled at 1000 Hz. Joint flexion at the knees and hips were determined from kinematic data collected with a 6-camera motion capture system (Vicon 460, Vicon Motion Systems, Oxford, UK) sampled at 100 Hz. Correlation analyses assessed whether peak impact force was associated with maximum knee joint flexion, immediately after impact, or skater experience.
Results: Decreasing peak impact force was significantly correlated with increasing knee flexion in four of six subjects (r ² -0.68, p ² 0.016, Figure 1). The two subjects with no correlation between knee flexion and impact force (r ³ -0.24, p ³ 0.25) had the least amount of experience (average 1.75 years vs. 7.25 years for the other four subjects). After excluding for a very experienced subject with new skates (circle symbol, right panel), peak impact force was found to significantly decrease as skater experience increased (r = -0.95, p <0.012). Hip flexion/extension behavior appears to be inconsistent between subjects.
Discussion: Our results suggest that more-experienced trick skaters tend to attenuate impact force. This appears to be accomplished partially by increasing knee flexion (or eccentric contraction of the quadriceps muscles) immediately after impact. Less-experienced skaters may be more concerned about maintaining balance, rather than refining their technique to (subconsciously) minimize impact force.

Figure 1: For the four most experienced subjects (hollow symbols), impact force significantly decreased with increasing knee flexion (see trend lines for each subject, left panel). If an experience subject with new equipment (circle) is omitted, then average impact force over all ten trials significantly decreases with increasing skater experience (right panel).