Neville Hogan, Ph.D. - US grants

When an engineer is attempting to design an artificial arm, a great many different data are needed concerning the forces and torques which the biological arm is put through each day. It is these "natural" forces and torques that the artificial arm must imitate. This research has designed an apparatus that willmake measurements on human arms and, thus, provide much of the necessary data for such designs. In addition, this apparatus is programmable, so that it can be used directly in the development of the software, i.e., computer programs, that will be used to control the artificial arm.

This fundamental engineering research should produce a new technology for cognitive rehabilitation. An intelligently controlled device will be fabricated for persons with cognitive impairments. This instrument will be able to analyze a patient's arm movements, distinguish between good and poor performance, then guide the person in making a more appropriate response. This device will be able to guide the person through unlearned and strongly guided motions to well- learned and non-guided motions. This new technology should be a valuable aid to persons who have had stroke, head trauma or other cognitive disabilities.

The broad goal of this application is to investigate the fundamental requirements for effective use of tools. The working hypothesis is that the central nervous system may take advantage of the apparent mechanical behavior of the neuro-muscular system (e.g., muscle elastic and viscous properties, skeletal geometry and inertia) to circumvent some of the major computational problems of controlling posture, movement and interactive tool-using behavior. The proposed research will study, kinematically-constrained motions, (e.g. opening a door) in which the hand may move only in certain directions. Experiments are proposed to evaluate the ability to perform a simple but highly-informative task- turning a crank (a handle mounted on a rod constrained to rotate about a pivot point). The experiments are to be augmented by and coordinated with theoretical studies. A detailed computer simulation of the upper extremity performing constrained motions will be developed. The work will also test a novel, biologically-motivated approach to controlling externally-powered arm prostheses which has been devised to enhance an amputee's ability to use tools. The controller makes a prosthesis mimic aspects of natural arm behavior: the response to external forces varies with co-activation of agonist and antagonist muscles while differential activation generates motion. The research method features an unique computer-controlled arm prosthesis. It can be programmed to mimic the behavior of any self- contained prothesis-existing or proposed. With the facility experiments can be performed which are not otherwise feasible with human subjects. Key aspects of the behavior of the (artificial) forearm can be manipulate directly-the way it responds to the human, the way it responds to the external world. Parallel studies of able-bodied subjects are planned to complement the experiments with amputees using the programmable prosthesis. Including amputees using powered arm prostheses in the study permits unique non-invasive perturbation experiments on human subjects. At the same time, a comparative study of able-bodied and amputee constrained- motion behavior will provide quantitative design specification for prosthesis control systems.

DESCRIPTION (Adapted from the Applicant's Abstract): The broad goal of this application is to investigate the fundamental requirements for effective use of tools. The working hypothesis is that the central nervous system may take advantage of the apparent mechanical behavior of the neuromuscular system (e.g. muscle elastic and viscous properties, skeletal geometry and inertia) to circumvent some of the major computational problems of controlling posture, movement and interactive tool-using behavior. The proposed research will study kinematically-constrained motions (e.g. wiping a surface), in which the hand may move only in certain directions. Experiments are proposed to evaluate the ability to perform several simple tasks: turning a crank (a handle mounted on a rod constrained to rotate about a pivot point); reaching to a visual target; pushing on a surface; and using a hand-held power drill. The experiments are to be augmented by, and coordinated with, theoretical studies. The work will also test a novel, biologically-motivated approach to controlling externally-powered arm prostheses which has been devised to enhance an amputee's ability to use tools. The control system makes a prosthesis mimic aspects of natural arm behavior; the response to external forces varies with co-activation of agonist and antagonist muscles, while differential activation generates motion. The research method features an unique computer-controlled arm prosthesis. It can be programmed to mimic the behavior of any self-contained prosthesis, existing or proposed. With this facility, experiments can be performed which are not otherwise feasible with human subjects. Key aspects of the behavior of the (artificial) forearm can be manipulated directly, the way it responds to the human, the way it responds to the external world. Parallel studies of able-bodied subjects are planned to complement the experiments with amputees using the programmable prosthesis. Including amputees using powered arm prostheses in the study permits unique non-invasive perturbation experiments of human subjects. At the same time, a comparative study of able-bodied and amputee constrained-motion behavior will provide quantitative design specifications for prosthesis control systems.

DESCRIPTION: (adapted from Investigator's abstract) Stroke is the leading cause of disability in the U.S. The economic burden of stroke was estimated to be $30 billion in 1993, equal to 3% of national health expenditures. Much of that cost is due to the highly labor-intensive nature of present rehabilitation practice which suggests that it may be possible to use robotics and information technology to improve the productivity of the health care delivery expert and at the same time improve a stroke victim's quality of recovery. The broad goal of this research project is to use robotics to study stroke recovery and improve neuro-rehabilitation treatments. Current research into recovery from brain injury posits activity-dependent plasticity underlying neuro-recovery. That motivated the specific aims of this project, to (1) test whether sensory-motor activity facilitates significant recovery of motor function in patients recovering from stroke and (2) test whether recovered motor performance of stroke patients exhibits characteristics associated with normal motor learning. If true, this would provide a basis to adapt mathematical learning theories into a quantitative theory of motor recovery. Using robotics and information technology for neuro-rehabilitation will provide, for the first time, objective control and quantification of the motor activity delivered to a patient as well as precise and reliable measurement of patients' motor behavior, thus enabling a rigorous test of these hypotheses. Briefly, patients with stroke will receive standard post-acute hospital care in a defined stroke rehabilitation setting. Patients in the experimental group will also be given robot-administered training in the form of sensory-motor manipulation of their impaired upper limb. Outcomes will be measured using standard clinical instruments as well as novel robot-based measures. A follow-on study will test for evidence of specific behavioral features that may be characteristic of motor learning by comparing the recovered motor behavior and motor learning ability of out-patients with age-matched normal subjects. It is expected that results from this study will provide an objective basis for maximizing the benefits of at least this kind of (robot-administered) therapy; and may lay the groundwork for a quantitative theory of motor recovery and possible further refinements of neurologic rehabilitation. In the case of alternative outcomes it is expected that this systematic approach combined with the quality of robot-based measurements will contribute to a scientific foundation for neurologic rehabilitation.

DESCRIPTION (Adapted from the Applicant's Abstract): Stroke is the leading cause of disability in the U.S. The economic burden of stroke was estimated to be $30 billion in 1993, equal to 3% of national health expenditures. Much of that cost is due to the highly labor-intensive nature of present rehabilitation practice which suggests that it may be possible to use robotics and information technology to improve the productivity of the health care delivery expert and at the same time improve a stroke victim's quality of recovery. The broad goal of this research project is to use robotics to: 1) study stroke recovery, and 2) optimize neuro-rehabilitation treatments. The working hypothesis is that there are at least two major aspects of neuro- recovery: 1) a process analogous to motor (re-)learning to compensate for damage to brain centers for coordination and control, and 2) a process analogous to recovery of strength and/or muscle tone. The Specific Aims of this study are to understand and distinguish between the effects of rehabilitation to enhance recovery of sensori-motor coordination and rehabilitation to restore muscle tone, strength and the ability to move against gravity. Briefly, patients will be given conventional therapy and in addition, different forms of robot-administered therapy: some patients will receive "placebo" therapy (in which the robot is inactive), others will receive sensori-motor training and others will receive progressive resistance exercise. Outcomes will be measured using conventional clinical instruments and also novel robot-based measures of coordination and muscle tone. Initial studies will focus on horizontal plane motions with the arm. Later studies will investigate motions against gravity, with and without robotic assistance. If the hypotheses are confirmed, it is expected that this study will provide an objective basis for: 1) maximizing the benefits of at least these two kinds of (robot-administered) therapy; 2) customizing this therapy to meet patients' specific needs; and 3) further refinements of robot neurologic rehabilitation. Should the hypotheses be falsified, it is expected that this systematic approach combined with the quality of robot-based measurements will contribute to the scientific foundations of neurologic rehabilitation.

DESCRIPTION (provided by applicant): Motor systems neuroscience is a branch of neuroscience rapidly gaining in importance. Whether one talks about clinical applications (such as using Deep Brain Stimulation as a treatment of Parkinson's disease or robotic rehabilitation devices for stroke patients) or basic science discoveries (recent findings of a surprising degree of plasticity in primary motor cortex), an understanding of how neural circuits control movement is critical to understanding how the brain works. At the same time, technological advances in bioengineering - such as the ability to record simultaneously from hundreds of neurons through multiple chronically implanted microelectrode arrays - have generated huge streams of data related to the neural control of movement. For these reasons, there has been a surge in theoretical (or computational) motor systems neuroscience research in an effort to codify our understanding of the global circuit principles that underlie motor learning and execution. Sophisticated mathematical theories and computational techniques are now routinely used to aid in the understanding of physiological data and behavioral findings. Given this rapid growth of theoretical motor systems neuroscience, there is concern of a growing divide between the computational scientists and the physiologists. In particular, continued progress depends on physiologists understanding the critical concepts the theorists are putting forward (regardless of whether they understand all the mathematical details) and on the theorists presenting their theories in ways that lead to testable consequences. The purpose of this satellite - an adjunct to the 2009 Neural Control of Movement conference -- is to foster better communication between theoreticians and experimentalists by grounding their interaction in the universal scientific language: experimental predictions. On four thematically-organized panels, key ideas in theoretical motor systems neuroscience (such as the synergy hypothesis and Bayesian optimality) will be discussed. The emphasis will be on determining the key empirical consequences of these ideas and discussing the likelihood of developing viable experimental strategies to test these predictions. The satellite is designed to cultivate group discussion with three discussants on every panel. PUBLIC HEALTH RELEVANCE: The field of motor systems neuroscience is concerned with understanding how the brain controls movement, so that we can better treat patients whose motor circuits have been compromised by disease or injury, such as stroke patients or patients with Parkinson's disease. Recently, there has been a surge of interest in theoretical (or computational) motor systems neuroscience to help organize large amounts of data with a few key mathematical principles. The purpose of this satellite is to foster better communication between the theoreticians and experimentalists working in the field by focusing on the key experimental predictions of a few important theoretical ideas.

From throwing a baseball to playing the piano to typing on keyboards, human beings are constantly learning new sensorimotor skills. During learning, synaptic connections in the brain must be modified to form a motor memory. Further, this modification seems both permanent and robust: a sensorimotor skill, once learned, tends to persist throughout the course of a lifetime regardless of its salience (recall the old adage of never forgetting how to ride a bike). Despite the importance of motor memories, their distinctive features, and their ubiquity in vertebrate behavior, little is known about the computational principles and mechanisms that subserve the acquisition of sensorimotor skills. This US-Canadian collaborative project takes an interdisciplinary approach aimed at elucidating neural mechanisms of motor memory formation and unifying -- under a common theoretical principle -- the findings of single-neuron recording studies with established behavioral results. The theory that is proposed makes the following testable prediction: as the level of behavioral expertise in a specific task increases, the neural representation for that skill becomes more selective. By selective, it is meant that a neuron significantly recruited during the performance of the skill tends, with practice, to specialize by firing only when that skill is performed (and not when related skills are performed).

Central to the theory is a geometric interpretation of "biologically plausible" sensorimotor neural networks, in which neurons are modeled as noisy signal processors and synaptic change is modeled as a noisy morphological process. Because of the high noise levels, it is shown that the system must be "hyperplastic" -- that is, the learning rate must be unusually high in order to compensate for the noise and operate at an acceptable performance level. Geometrically, the solution for a skill can be represented as a manifold in the weight space of the network. To learn multiple skills, a network configuration must be attained such that the solution manifolds intersect. To learn multiple skills without noise leading to destructive interference, the network must arrive at a point where the intersecting solution manifolds are orthogonal. With this principle of orthogonality, the neurophysiological predictions described above can be explicitly formulated. These predictions will be tested with an experimental method -- involving floating microelectrode arrays and antidromic stimulation -- that enables the identifiably same neuron to be recorded from for multiple days/weeks, while a behaving animal learns a task. Finally, psychophysical predictions of the theory will also be tested.

This project is jointly funded by Collaborative Research in Computational Neuroscience and the OISE Americas program. A companion project is being funded by the Canadian Institutes of Health Research.

This EArly-concept Grant for Exploratory Research (EAGER) collaborative research project is between an expert in robotics and control theory and an expert in experimental and computational motor neuroscience. It bridges cognitive science, experimental psychology and control engineering. The intellectual premise of the work is that a quantitative theory of human cognition may be built on top of limiting cases of human motor function. This premise lays the foundation for the development of a comprehensive quantitative theory of control-relevant cognition. The result will be an invaluable tool for the human-friendly design of complex motion control systems. Control strategies based on these fundamental objects would be more intuitively understandable by human operators, including prediction of impending failure. Extension of the results beyond motion control provide a new class of knowledge-processing systems capable of more natural interactions with humans.

The objective of this project is to articulate and test a quantitative, control-relevant theory of human cognition, to address a growing divide between cognitive science and control theory. The core hypothesis is that cognitive functions emerged from and are constrained by neural structures used for motor control. Complex motor actions are composed from a limited "library" of dynamic primitives, defined as attractors (e.g. fixed points, limit cycles, etc.). The project postulates that a similar composition of dynamic primitives underlies cognitive processes and that quantitative details may be obtained by re-purposing dynamic primitives found in motor behavior, especially in the manipulation of complex objects such as tools where the link between motor and cognitive function may be strongest. The project is based on a novel series of experiments: A data series is generated by various human participants physically manipulating a complex dynamic object. Alternative data sets are generated by computer simulation of movements to the same targets that minimize mean-squared applied force. Random fluctuations generated by low-pass filtered zero-mean Gaussian white noise and of magnitude comparable to the fluctuations in human performance are added to the simulated force and motion time-series. Without being told the origin, a second set of subjects are presented with the results as evolving abstract time-series and asked to predict their outcome. Subsequently, they are asked to generate a control input for the abstract system, based on their experience, to accomplish a specified task. According to the hypothesis, the subjects will more successfully predict the outcome of human-controlled systems than the synthetic systems, and will generate control inputs that more closely match the human-controlled system inputs.

There are many arenas where humans outperform machines. For example, when coordinated interaction with the physical environment is needed, humans (and animals) vastly out-perform modern robots. This occurs despite the biological systems having far slower 'hardware' and 'wetware' and much greater complexity than even the most modern robots. This research project seeks to understand the role of complexity in human sensory and motor performance. Human walking under challenging balance conditions will be studied and the use of canes to enhance stability will be included. The investigators emphasis on learning to balance in challenging environments should lead to new rehabilitation therapies (with or without robotic assistance) to aid recovery of balance and walking (e.g., after stroke). The researchers will create educational units suitable for online presentation to K-12 students and will devise exhibits based on their research for the Museum of Science in Boston.

The central hypothesis to be tested in this project is that complex movements involving physical interaction with objects are organized as a hierarchy formed of modules or primitives. Experiments will study how unimpaired humans learn to walk on narrow beams. Beams of different roundness will vary the challenge. Hand-held canes will alter the available support (like training wheels on a child's bicycle). Computer simulations combined with machine learning will study the benefits and drawbacks of organization as a hierarchy. New mathematical tools will be developed and tested to see if they enable insightful description of human performance in challenging conditions. The research involves a multinational collaboration among scientists from the U.S., Israel, and Germany, each with complementary expertise. The bridge between experimental and theoretical work and the diverse Principal Investigators will help to attract women into the traditionally male-dominated fields of computational neuroscience, robotics and control engineering. Companion projects are being funded by the Federal Ministry of Education and Research, Germany (BMBF) and the US-Israel Binational Science Foundation (BSF).

Despite vastly slower "hardware" and "wetware," human dexterity vastly out-performs modern robots. This project studies apparently-simple tasks - managing the kinematic constraint on hand motion required to open a door; and dealing with the dynamic complexity of liquid sloshing in a cup of coffee - that profoundly challenge robots but humans perform with ease. The key idea is that humans manage skillful physical interaction with these objects by exploiting clever combinations of primitive dynamic actions that do not require continuous intervention. A novel theory to describe the effectiveness of this approach is developed and tested by experiments with human subjects. The theory is applied to transfer comparable skill to robots, despite manifestly different hardware. If successful, these robots will be more capable, more comprehensible, and more collaborative partners with humans.

The central experimental challenge is to determine the essential strategy underlying humans' remarkable competence in physical interaction tasks. Three hypotheses reflecting major themes in contemporary motor neuroscience are tested: Humans 1) develop models of object dynamics sufficient to pre-compute and execute required hand motions (similar to modern robot programming); 2) choose forces and motions to minimize muscular effort (similar to optimizing efficiency); or 3) exploit dynamic primitives to robustly achieve satisficing (good-enough) performance. The theoretical challenge is to formulate a coherent account combining the information-processing of brains (or computers) with the "energy-processing" of physical objects and their interactions. Classical equivalent circuit theory is re-purposed to define a neo-classical equivalent network theory, combining dynamic motion primitives with mechanical impedances (interactive dynamics). Mechanical impedances enjoy a key property, compositionality, that overcomes the curse of dimensionality.

Human sensorimotor capabilities vastly out-perform those of modern robots. Prior research suggests that humans achieve skillful movement by exploiting combinations of "dynamic primitives", which are robust building blocks of coordination that simplify control. Interaction with complex objects - such as spreading a tablecloth - requires prediction which, in turn, requires mental representations of the objects and environment. This project explores the extent to which such mental models may also take the form of dynamic primitives. The project team will perform fundamental research exploring: how humans learn to manipulate a complex flexible object through the composition of dynamic primitives; the impact of explicit instruction on the acquisition of the mental models; and whether a primitive-based control structure to be implemented in a robot can achieve skillful manipulation of complex objects and fluid interactions between humans and robot. This project serves the national interest because the resulting understanding of human sensorimotor control and robot control methods may result in improved efficacy of robot-assisted physical rehabilitation after neuromotor injuries such as stroke, where safe physical cooperation with humans is fundamental and mutual learning is of paramount importance. The project will involve an educational component that provides engineering and research methods training to graduate and undergraduate students, as well as STEM outreach to individuals of all ages at the Museum of Science in Boston. Additional efforts will be made to attract and retain women into careers in science and engineering.

This research investigates a biomimetic approach to robot control that promise improved human-robot physical interaction during co-manipulation of flexible objects with complex continuum dynamics. Methods include skill acquisition experiments involving expert and novice human subjects to test how mental representations are formed and the extent to which they may be shaped by explicit instruction during training; a theoretical study of how motor learning may be facilitated using mental models based on dynamic primitives vs. lumped mechanical properties; and a study of human-robot co-manipulation of flexible objects wherein dynamic primitives provide the basis not only for robotic control, but also for communication and mutual learning between the human and robot partners. This research promises new insights into the manipulation of complex objects and tools, an area where humans still outperform machines.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

Project Summary A large body of research has led to statistical models showing how movement velocity is encoded in the motor cortex. These models have been validated by the control of neural prosthetics which restore natural arm and hand movement to paralyzed individuals. However, forces also need to be controlled in harmony with motion when interacting with objects. While many studies have examined force and motion separately, research has rarely focused on how the motor system coordinates both together. The simultaneous variation of force and motion is incorporated in the definition of impedance. Our current neural models do not describe impedance encoding, which poses severe limitations on our understanding of the control of object interaction, an important aspect of human behavior. The proposed research will develop new models of motor cortical impedance encoding during object interaction. Using these new models to decode ongoing impedance signaling, we will substantiate an advanced theory of impedance control used by the motor system to produce accurate object displacement in response to the forces applied by the hand. This research bridges the expertise of Dr. Schwartz in neurophysiology and of Dr. Hogan in robot control. Monkey subjects will perform tasks with real and virtual tools that naturally encourage the use of impedance control. We will record the activity of motor cortical neurons during these tasks and develop new mathematical models to describe the relation between neural activity and force, motion and impedance. Results from electromyography recordings, joint angle measurements and torque calculations, together with the neural models, will be used to better understand how impedance is regulated at the level of muscles and joints and how the impedance of the hand is signaled during object interaction. This work promises to extend our understanding of the neural control principles governing the way we use our arms and hands to interact with our surroundings. These principles can be used to build new theories of the cognitive processes used to predict and effect changes in the world around us. At the same time, elucidation of the neural and mechanical details of forceful interaction will lead to new rehabilitative and neural prosthetic approaches to paralysis.