clockwise from bottom: W. Thomas Edwards, PhD, director, Rehabilitation Engineering and Rehabilitation Engineering Analysis Laboratory (REAL) at Kessler Medical Rehabilitation Research and Education Corporation (KMRREC), associate professor, physical medicine and rehabilitation (PM&R), UMDNJ-New Jersey Medical School (NJMS); Lisa Simone, PhD, Senior Research Scientist (REAL); Venkata Gade, Research Engineer (REAL); and David Tung, MD, Resident, PM&R, NJMS.

Movement research: from lab to the real world by W. Thomas Edwards

Stroke and traumatic brain injury (TBI) can adversely affect the neuromuscular system and result in weakness, balance deficits and spasticity. Through studies of motion and function, we have learned that potentially important changes can be both subtle and inconsistent. Several current projects in the Rehabilitation Engineering Analysis Laboratory (REAL) at KMRREC are developing new techniques to address these needs.

One area of interest is to determine the effect of brain injury on postural control. The risk of falling and fall-related injuries increases following TBI or stroke, with about 40% of these individuals requiring treatment. Balance requires the interaction of various sensory systems, proprioception, and muscle contraction, with the kinetic response of the body segments. Healthy individuals commonly use two strategies for maintaining balance. These are the ankle strategy, in which rotations occur at the ankle with little rotation at the other joints, and the hip strategy, in which rotations at the hip provide stability. These strategies are reflected in the torque provided at the ankle, knee and hip and produce distinct patterns of motion. Although the human postural control mechanism has been studied extensively, many questions remain. We feel that methods to better identify these strategies are needed. These methods can be based on the measurement of patterns of motion.

A pattern of motion is described by an associated frequency and a set ratio of relative rotations of the joints during sway oscillations. We have employed an approach combining mathematical modeling and experimental measurements to quantify the patterns of motion in response to dynamic motion stimuli, a moving platform. Considering sagittal motions, the mathematical models simulate the interactions of the ankle, knee and hip with the control of the mass of the body. Balance is maintained by varying the torque and stiffness at the joints in response to disturbances. An interesting finding from the modeling studies is that the body can be tuned to control the relative rotation at each joint. By adjusting the torques at the ankle, knee and hip, a pattern of motion can be selected which provides balance with the least effort.

To measure these responses in our lab, we record ground reaction forces, muscle activity and body motions. A NeuroCom Research Balance Platform is used to apply motions and measure forces at the ground. The NeuroCom Research Platform system is a programmable dynamic platform with two independent AMTI force-plates. This research platform has the capability to assess balance using either clinically based motion protocols or more general investigator selected procedures. The platform motion is triggered using the REAL’s VICON motion analysis system. Together these systems provide a complete description of the balance responses under different dynamic test conditions in response to the disturbances at the ground. Studies using low frequency platform translates have shown that subjects can select different strategies to improve stability at each frequency.

A challenge is now to merge our modeling experience with our experimental observations. This requires the development of techniques that can identify the patterns of motion from the experimental data. The experiments using the moving platform are designed to elicit patterns of motion involving the hips and upper body, in addition to rotation at the ankles. We have learned in collaboration with Haim Baruh, PhD, a professor at Rutgers University, that engineering techniques have been developed to identify the properties and behavior of structures that could be applied here. A technique called modal analysis can successfully locate defects (deficits) or refine the description of properties (such as stiffness and damping) of a structure. For application to postural control, this and other system identification methods based on the assessment of the patterns of motion do not require knowledge of the underlying physiological processes. Evaluation of the patterns of motion will provide a quantitative measure of behavior, such as that involving an “ankle” or “hip” strategy. These studies have already led to some very interesting questions. For example, we found that for some individuals, the change in torque as the body rotates over the ankle is only one-third of the amount theoretically needed to stand. It is not clear how these people are able to maintain their balance. We believe that coordinated motions of the upper body could provide an explanation. Findings from these projects were presented recently at the Gait and Clinical Movement Analysis Society and at the Biomedical Engineering Society.

Laboratory based studies such as these postural control experiments have both advantages and disadvantages. On one hand, the laboratory offers a controlled environment with large scale, precision instruments; on the other, large fixed systems limit access to many groups of subjects. For balance studies, REAL investigators can reach other populations, such as individuals in a nursing home, for short term measurements using mobile force plates. However, investigators at KMRREC identified a need to monitor patients over longer periods of time in more natural settings.

Because clinical measures don’t always tell us how well someone can interact in the home environment, Lisa Simone, PhD, a senior research scientist at REAL, created a low-cost wearable device to measure hand function outside the clinical setting. She describes the need for this device: “When I began working with individuals recovering from brain injury, I was astounded by how intricate our motor control systems must be to perform even simple tasks with our hands, and how intrusive the effects of brain injury are on the ability to perform basic daily activities. One of our participants who had a stroke told me he just wanted to hold the club and play golf again; another, to turn the key in her car’s ignition to drive to work. Our progress in addressing basic hand function is frustratingly slow, so we decided to jump beyond the clinical setting and focus on measuring and understanding how individuals use their hands and fingers in their homes and communities. Because every individual has different goals and tasks that he considers critical to get through a typical day, it behooves us to better understand how treatment affects these goals, and optimize our rehabilitation efforts accordingly.”

Dr. Simone’s Shadow Monitor is a lightweight wireless device that records finger postures over a 24-hour period while individuals participate in their daily activities. In a collaborative effort with Derek Kamper, PhD, of the Rehabilitation Institute of Chicago, we recently achieved a significant milestone to “go wireless” with the monitor. Finger posture data can now be stored in the monitor or transmitted wirelessly to a nearby computer. In this way, the wearer does not need to be tethered to a computer, and can move about the home or in the community while data is collected. To achieve long battery life in the field (several days), we integrated a newer wireless protocol for personal area networks (IEEE 802.15.4) designed specifically for low power, low data rate monitoring applications like the Shadow Monitor.

Recently, the Shadow Monitor team (which also includes bioengineer Nappi Sundarrajan, MS, and Elie Elovic, MD) completed a research study designed to evaluate usefulness of the hand monitor in individuals with brain injury (funding provided by the Foundation of UMDNJ). The monitor can clearly differentiate between different hand activities — like writing with a pen or picking up large objects — for healthy individuals; we found that the measured quality of the finger movements can degrade significantly with motor dysfunction. Our challenge is to correlate specific therapies with quality of movement measured with the device in order to predict which may be ideal for optimal functional recovery.

Initial results show that the monitor is capable of measuring significant differences in hand performance based on ability level. We compared different measures of joint range of motion (ROM) in order to identify a method that reflects an individual’s ability to perform daily tasks. Both active ROM (the individual moves all joints maximally) and functional ROM (maximum joint movement during specific functional activities) have been proposed as useful measures of function. Comparing individuals with brain injury and healthy individuals, we found that active ROM is a useful predictor of functional ability, although functional ROM did not show a significant difference between the groups. While this may appear misleading, it simply forces us to reevaluate just what is meant by “function” and how best to measure it effectively.

Another significant finding in Dr. Simone’s study is that joint speed and acceleration while performing activities is significantly related to ability level, while stationary hand position and joint angles are less so. Individuals with brain injury open and close their fingers nearly three times slower than healthy individuals, even though the time to complete individual tasks is only slightly longer. This may provide a more sensitive picture of hand function with disability level beyond what can be assessed using traditional range of motion measurements. Dr. Simone will be continuing her research and teaching design at New Jersey Institute of Technology in Newark after she joins the faculty in the Department of Biomedical Engineering this winter.

Improving the assessment of function in the laboratory and in other settings is a challenging rehabilitation engineering objective. We are working through these projects to lay a foundation for clinical evaluation of patients and to identify novel rehabilitation strategies for those following stroke or TBI.

W. Thomas Edwards, PhD, is an associate professor in the Department of Physical Medicine and Rehabilitation at UMDNJ-New Jersey Medical School, and director of Rehabilitation Engineering and the Rehabilitation Engineering Analysis Laboratory at KMRREC. He received his PhD from the Massachusetts Institute of Technology. He has served on the editorial boards of several journals, most recently the Journal of NeuroEngineering and Rehabilitation. Dr. Edwards also reviews for the NIH Musculoskeletal Rehabilitation Sciences Study Section and the NIDRR Health and Function review panel. His research is in musculoskeletal and spine biomechanics with a current focus on postural control.