Jonathan B. Dingwell, PhD - US grants

Every day, humans interact with objects that challenge their ability to control movement stability, such as carrying a cup of hot coffee. For patients with movement disorders, such as those associated with basal ganglia or cerebellar disease, such tasks can significantly impact their ability to perform everyday tasks. However, surprisingly little is known about how the central nervous system (CNS) controls movement when interacting with dynamically complex objects. The CNS uses internal representations of object mechanical properties to manipulate simple rigid objects, and can also learn internal models of complex external force fields for making reaching movements. It was hypothesized that subjects learn similar internal representations of the dynamical properties of complex objects. Preliminary experiments from our lab indicate that healthy subjects adapt their movements to maintain stability of non-rigid objects during a goal-directed reaching task. The proposed research will first endeavor to determine if this adaptation occurs as the result of the development of an internal model of object dynamics, and second, to characterize those aspects of adaptation that are specifically related to maintaining movement stability when manipulating complex objects. Experiments will be conducted using a robotic manipulandum that allows the intrinsic dynamical properties of complex "virtual objects" to be precisely defined. In the first experiment, subjects will be trained to make point-to-point reaching movements while manipulating a dynamically complex virtual object. A comparison of figural errors (which quantify differences in kinematic shape profiles) between movements made with dynamic and rigid objects will be used to determine if subjects are learning and internal model of the object dynamics, or are simply increasing overall limb stiffness or memorizing the specific patterns of forces imposed by the manipulandum. In the second experiment, subjects will make continuous rhythmic movements with the dynamic object. Methods from nonlinear dynamics will be used to quantify movement dimensionality (correlation dimensions) and the sensitivity of the neuromuscular control system to internally-generated local perturbations (Lyapunov exponents). It is hypothesized that the dimensionality of movement is determined by the intrinsic mechanics of the arm+object, and will therefore not change significantly as a function of training, but that the central nervous system adapts its control strategy to make movements less sensitive to local perturbations. It is anticipated that the results of these experiments will lead to a better understanding of the neuromuscular control processes underlying object manipulation and will eventually help provide clearer direction for developing functionally realistic virtual interactions for motor rehabilitation.

DESCRIPTION (provided by applicant): Repetitive strain injuries (RSI) affect thousands of people and cost the US economy more than $14 billion each year. It has long been believed that improper postures or movements made during repetitive tasks increase the risks of developing RSI. Muscle fatigue may be an important intermediary factor in this process, since muscle fatigue can induce changes in coordination that generate improper movements, which may in turn increase the risk of RSI over time. The purpose of this R21 application is to develop and test the ability of new methods to track the changes that occur in both muscle function and coordination during fatiguing repetitive movements. A device will be constructed to simulate an upper extremity repetitive task known to induce changes in coordination after fatigue. Appropriate analytical tools for tracking fatigue from observed changes in coordination will also be developed by extending existing nonlinear dynamics algorithms developed for tracking damage accumulation in mechanical systems, Because this approach tracks distortions in appropriately reconstructed state spaces, it can provide valid measures of the underlying (hidden) damage dynamics without the need for detailed physics-based mathematical models of either the system or damage dynamics. Currently available algorithms will be modified to account for the most prominent differences between mechanical and biological systems: noise, multiple time scale dynamics, and non-monotonic damage dynamics (i.e. biological adaptability). Finally, these and more traditional measures will be applied to explore the time courses of changes in muscle function and motor coordination that occur during the repetitive work task. 30 healthy subjects will perform the task until voluntary exhaustion under three conditions: more restricted, less restricted, and less restricted at elevated work height. It is hypothesized that (1) changes in local muscle fatigue will precede changes in muscle coordination, which will in turn precede overt changes in kinematics, (2) this sequence of events will be delayed in the less restricted condition, (3) these changes will occur more rapidly in the elevated work height condition, and (4) the nonlinear tracking approaches will reveal subtle changes in coordination that reflect underlying (hidden) changes in muscle fatigue state. This project will generate new insights into the nature and time course of the biomechanical and neural adaptations that occur during repetitive tasks and will provide the necessary foundation for developing improved diagnostic techniques to identify early-onset (pre-clinical) RSI in future work. It is hoped that these efforts will one day help reduce the tremendous personal and monetary costs associated with these injuries.

This interdisciplinary research program is focused on the development and application of a new class of nonlinear dynamical system inspired by the study of repeated precision human movements. The fundamental qualitative dynamics of these systems will be examined. Our approach is based on a novel definition of movement tasks in terms of goal functions, which encode the interaction between the body and the environment needed for perfect execution. The resulting task dynamical systems model the trial-to-trial performance of precision goal-directed actions, such as, for example, the repeated throwing of a ball at a target or the hammering of a nail. Using these models, one can examine how performance arises from the interaction between the geometry of task-specific goal equivalent manifolds, passive sensitivity, control and active stability, and intrinsic noise.

New concepts and methods will be developed for modeling and characterizing trial-to-trial variability in a range of different tasks. The resulting theoretical models will be used to study the mechanisms of variability generation in repeated movements, to determine the simplest model features that are capable of exhibiting observed phenomena, and to examine the dynamical implications of specific control assumptions. The theoretical results will be used to develop new experimental methods and to make experimentally testable qualitative predictions. The resulting conceptual framework and analysis methods will enable researchers to, for the first time, untangle the passive mechanical aspects of movement from the perceptual and neurological aspects believed to be related to active control.

This project involves collaboration between researchers in nonlinear dynamics, movement science (kinesiology), and robotics. In addition to addressing fundamental questions in movement science, our effort will lead to: new noninvasive approaches to studying, monitoring, and diagnosing neurological movement disorders; monitoring progress in physical therapy after surgery or injury; and characterizing motor learning and performance in subjects involved in sports and other precision tasks. In the area of technology, our work will find application in the design of biologically-inspired precision machines and advanced man-machine interfaces needed for such applications as remote
tele-surgery. Such machines will be designed to exploit task redundancy and passive stability properties so that they inherently respond to changes in their operating environment, or internal changes caused, for example, by component wear. This, in turn, could lead to improved repeatability, reliability and service life, even for entirely open loop devices.

[unreadable] DESCRIPTION (provided by applicant): Work-related musculoskeletal injuries have been linked to movement repetition, improper postures, and muscle fatigue. Muscle fatigue and/or the muscle imbalances that result from fatigue may be important as direct causes and/or intermediary factors in injury development. To develop appropriate prescriptions for preventing injuries and effective strategies for rehabilitation, it is critical to understand how fatigue muscle and muscle imbalances affect the control of goal-directed movements. [unreadable] Thirty healthy young adults will perform an upper extremity low-load continuous work task similar to sawing for 5 minutes in time with a metronome. They will then perform either of two fatigue protocols to induce either widespread fatigue (more sawing with increased resistance) or localized muscle fatigue (repetitive shoulder flexion). They will then perform another 5 minutes of low-load sawing. Between each segment of the protocol, rates of perceived exertion and maximum voluntary contractions (MVCs) will be recorded. Kinematics, handle forces, and muscle activity (EMG) will be recorded continuously during all sawing trials. Proper movement timing is critical to many repetitive tasks. The PIs will apply novel methods to decompose the variability in the primary spatial-temporal task variables into new variables that directly affect achieving the task goal and those that do not. They will quantify both the variability and cycle-to-cycle temporal correlation structure of each resulting time series to determine how humans control movement timing during redundant goal-directed movements and how this control is altered with either widespread or localized muscle fatigue. [unreadable] Controlling movement stability during repetitive tasks is also critical for minimizing risk of injury. However, how muscle fatigue and/or muscle imbalances affect the control of movement stability is not well understood. The PIs will apply innovative analysis methods, developed by their lab, to directly quantify each subject's innate sensitivity to small perturbations during the continuous sawing task trials. These analyses will allow them to determine how both widespread vs. localized muscle fatigue affect the control of movement stability. [unreadable] This R03 project will yield vital new insights into how muscle fatigue affects task performance, about the neuromuscular and biomechanical strategies humans use to achieve this performance, and about the underlying control policies subjects adopt. Answering these fundamental questions is critical for understanding the mechanisms of musculoskeletal injury development and for developing effective rehabilitation strategies for treating patients with these injuries. Upper extremity musculoskeletal injuries (not including low back pain) are a significant and costly health care problem affecting over 375,000 people in the work-place each year. Muscle fatigue and muscle imbalances are significant contributors to these injuries. Therefore, it is important to better understand how muscle fatigue and muscle imbalances caused by localized muscle fatigue affect how humans control the timing and dynamic stability of repetitive movements. [unreadable] [unreadable] [unreadable]

[unreadable] DESCRIPTION (provided by applicant): Many disabilities significantly disrupt walking, including neurological, muscular or orthopedic disorders, and normal aging. For example, as many as 12 million elderly over age 65 and 60% of lower extremity amputees fall each year. The total costs of all fall-related injuries could reach $43.8 billion by 2020. Identifying those at greatest risk of falling so proper interventions can be applied is critical to reducing these numbers. Most falls occur while people are walking. Therefore, the goal of this project is to develop appropriate tools to quantify dynamic stability during walking so we can solve this momentous clinical problem. In mechanics, stability is defined by how a system responds to perturbations. Global Stability defines the set of all perturbations a system can respond to without "falling over". Global stability in humans can be tested by imposing large perturbations like slips or trips. Local Stability defines how a system responds to very small perturbations. Our lab has developed novel approaches to quantifying local dynamic stability of walking and used these to validate several intuitive clinical observations regarding strategies patients use to maintain local stability during normal (i.e., unperturbed) walking. Our ultimate goal is to develop valid methods to predict falls without having to directly cause falls. Doing this will require determining if and how local stability is related to global stability. This is a very difficult problem because there is no theoretical guarantee that local stability will predict global stability and because the precise mathematical definitions of these quantities, derived for deterministic systems, are not easily applied to noisy biological systems. For this Exploratory / Developmental R21 project, we will first derive and validate a novel set of quantitative measures of dynamic stability that specifically account for stochastic "pseudo-periodic" motions and are thus appropriate for analyzing human walking data. Second, we will validate our stability measures using a novel biomechanical model designed specifically to analyze walking stability. Our dynamic walking model will incorporate sufficient muscle activation for forward propulsion, and bio-mimetic state feedback control with neuronal noise and physiological time delays to ensure lateral stability. We will conduct similar experiments in both the model and in healthy humans to determine how small-to-moderate perturbations affect local walking stability and how large perturbations affect global walking stability. Together, these efforts will tell us if appropriately defined measures of local stability, obtained during unperturbed or minimally perturbed walking, can predict actual risk of falling when our model and/or human subjects experience large perturbations. If so, the tools developed in this project could potentially significantly improve our ability to predict, and thereby prevent, falls in patients with locomotor disorders. These tools will also provide a coherent platform for determining the biomechanical and neurophysiological mechanisms humans use to prevent falls and for evaluating the efficacy of different therapeutic interventions intended to help augment these mechanisms. PUBLIC HEALTH RELEVANCE: Falls and the injuries that result from falls are a significant health care problem for the elderly and for patients with a wide range of walking disorders, including stroke, amputation, and many others. Finding ways of accurately predicting and preventing these falls will significantly extend and improve the lives of these patients. The proposed work will apply novel engineering concepts to directly quantify dynamic stability during walking to address this critical issue. [unreadable] [unreadable] [unreadable] [unreadable]

DESCRIPTION (provided by applicant): Walking is an extremely important and common daily activity. Many locomotor impairments increase people's risk of falling. The total costs of all fall-related injuries may reach $43.8 billion by 2020. As many as 60% of patients with lower extremity amputation fall each year. Falls are especially problematic for young patients with traumatic amputation, who fall slightly more than older patients. Most people fall while they are walking. Also, very limited scientific evidence exists to guide design of interventions to improve walking function in patients with amputation. Thus, there is a clear need to better understand how patients with lower limb amputation respond to ecologically relevant perturbations, to identify the biomechanical and neuromuscular strategies these patients use to recover balance after being perturbed, and to develop effective evidence based treatment strategies to help these patients improve their walking stability. Our lab has developed novel engineering approaches to measure walking stability that directly quantify how humans respond to small perturbations. The primary goal of this study is to develop interventions to help prevent falls. This requires intervening before the fall itself occurs. While falls themselves are very elusive events, significant stumbles are very common. In the elderly, stumbling or tripping causes more than half of all falls. Therefore, stumbling is one of the primary precursors to falling. Stumbles often lead to fear of falling, excessive caution, and decreased physical activity. Surprisingly, however, no study has quantified stumbling responses in patients with lower limb amputation. For this project, we will first determine how patients with trans-tibial amputation respond to small, continuous pseudo-random visual or mechanical perturbations, similar to those they might experience walking outdoors over uneven terrain or in crowded public places. We will also directly test the common clinical assumption that these patients rely more heavily on vision because of their loss of distal somatosensory feedback. Second, we will determine how patients with trans-tibial amputation respond to large discrete mechanical perturbations during walking, such as they might experience when tripping over a curb or stepping in a pothole. From these data, we will identify specific biomechanical and neuromuscular strategies amputees use to recover balance after they stumble. Finally, we will determine if targeted virtual reality based gait training is more successful than conventional therapy for improving walking stability in patients with trans-tibial amputation. A fully immersive virtual environment will allow us to apply highly controlled and ecologically relevant perturbations, which we anticipate will generalize more readily to real world walking. This study will apply novel experimental and rigorous analytical approaches to significantly improve our understanding of how patients with amputation respond to perturbations. We will translate this knowledge into clinical practice by developing rehabilitation interventions based on our scientific findings. Finally, this work will provide a scientific basis for developing better interventions to improve walking function in populations with other walking related impairments. PUBLIC HEALTH RELEVANCE: Falls and the injuries that result from falls are a significant health care problem for the thousands of patients who undergo lower limb amputation every year, as well as for millions of elderly and patients with other locomotor impairments. Determining how these patients use their available sensory and motor resources to regulate stability during walking and developing effective evidence based treatment strategies to improve their walking stability will significantly extend and improve the lives of these patients. The proposed work will apply novel state-of-the-art experimental and analytical approaches to directly address these critical issues.

? DESCRIPTION (provided by applicant): Falls are very common and extremely dangerous events for the elderly, constituting the leading cause of injury and incurring costs likely to exceed $68 Billion by 2020. Falls cause more than 95% of hip fractures and lateral falls (to the side) in particular contribute to 76% of all hip fractures. Most falls occur while walking and humans are inherently more unstable laterally when walking. Also, very limited scientific evidence exists to guide de- sign of interventions to improve walking function in the elderly. Thus, there is a clear need to identify the control strategies elderly use to maintain lateral balance while walking and to develop effective, evidence-based treatment strategies to improve lateral balance control in high fall risk elderly to reduce their risk of falls. The primary goal f this study is to develop interventions to help prevent falls. This requires intervening before fall occur. As people age, multiple physiological changes increase intrinsic physiological (neuromuscular) noise and decrease control authority (the ability to effectively regulate movements). Either or both of these can increase walking variability, which may contribute to falls. However, not all variability is detrimental. Increasing variability can sometimes even facilitate adaptability and improve recovery in locomotor rehabilitation. We recently developed novel computational control theory models that separate physiological noise from control authority to identify how walking humans exploit redundancy to regulate variability in the sagittal plane. We have now extended this work to determine how humans control lateral stepping movements in the frontal plane. Real-world walking tasks require humans both to be able to respond to changing task goals and also to choose effective strategies. Here, Aim 1 will determine how elderly with Low Fall Risk or High Fall Risk respond to externally imposed challenges (enforced step width and/or lateral perturbations). We will integrate our theoretical framework with computational models and experiments to differentiate effects of control from those of variability. Separately, Aim 2 will determine how elderly make (cognitive) internal choices to either avoid risk or fortify themselves against potential risk. We will again integrate experiments, models, and analyses to identify how Low and High Fall Risk elderly choose different risk-sensitive strategies. Aim 3 will determine if a targeted virtual reality based intervention that challenges people to both respond to imposed changes in lateral position, and also to choose effective strategies for doing so, can improve lateral stepping control and walking balance in High Fall Risk elderly. In a randomized, active control treatment trial, we will compare pre- and post-intervention changes in walking ability both with and without lateral perturbations, performance in a novel real-world-like navigation task, and established clinical assessments of walking and balance function. This study will apply novel experimental and rigorous computational and analytical approaches to greatly improve our understanding of how elderly individuals walk. We will translate this knowledge into clinical practice by implementing novel VR-based interventions that promise to improve walking function in high fall risk elderly.

Falls are the leading cause of fatal and nonfatal injuries in older adults (age ? 65), yielding extremely high injury rates & severity. Older adults experience significant declines in cognitive function (esp. executive function) that affect their motor performance and contribute to increased fall risk. This study will determine how differences in functional capacity (strength), functional walking ability, and particularly executive function (EF) affect the mechanisms both young and older adults use to regulate their stepping behavior during complex walking tasks. In two separate experiments, we will use state-of-the-art virtual reality technologies to manipulate the cogni- tive demands of a walking task in both young and older adults. We will carefully measure both physical and cognitive function in our subjects. We will then directly and systematically manipulate the cognitive demands of the walking tasks by introducing targeted competing task goals subjects must try to accomplish. This design will allow us, for the first time, to very carefully and precisely directly relate age-related changes/differences in sub- jects' underlying physical and cognitive ability to differences in their performance on the walking tasks. Specifically, we will quantify how these adults regulate ?Goal-Relevant? stepping errors with respect to differ- ent specified goals of the walking task. In Aim 1, we will determine how young and older healthy adults adapt their stepping strategies in the presence of spatial risk. Here, we will introduce a penalty and reward landscape to the treadmill space. Higher rewards (i.e. points) will be available closer to the penalty regions, creating conflict between increased spatial risk and increased reward. In Aim 2, we will determine how healthy and fall-prone older adults modify their stepping strategies to accommodate changing task goals that directly mimic real-world situations. These changing task goals will oblige older adults to make real-time cognitive decisions about where to step that have conflicting risks and rewards. In Aim 3, we will use data collected in the experiments of Aims 1 & 2 to determine how young and older healthy adults adapt their stride-to-stride anterior-posterior ground reaction forces to achieve the stepping strategies elicited. We hypothesize humans will actively exploit redun- dancies in the Force Impulse-Momentum Principle at each individual stride. Across all 3 Specific Aims, we will quantify how differences in physical and cognitive function directly affect the flexibility with which young and older adults can alter how they regulate their stepping movements from each step to the next. This R21 proposal is (scientifically) 'high risk' as it will be the first to systematically tie specific aspects of human decision making (executive function: EF) directly to the biomechanical outcomes of goal-directed move- ment (walking). This work will also uniquely identify how age-related declines in EF alter these strategies. How- ever, if successful, this work has great potential to be `high reward' because it will establish direct causal links between EF capacity and the control/regulation of walking movements that can be exploited to design evidence- based interventions to reduce fall risk in the elderly and/or other at-risk populations.