John W. Krakauer, M.D. - US grants

The proposed award is designed to develop the candidate's clinical research skills to prepare him for a career as an independent investigator in the application of motor psychophysics and functional brain imaging to the study of neurological disease. Research Plan: Stroke, Huntington's disease (HD), and Idiopathic Torsion Dystonia (ITD) are diseases that rob people of motor function in the prime of life. Many of the measures of motor performance and functional status commonly used in clinical trials and rehabilitation suffer from subjectivity and lack of scientific validation. The first goal of the proposed study is to characterize and quantify the motor deficit in these diseases using methods previously developed in the study of arm movements in normal subjects. In particular, we will emphasize the importance of examining motor learning abnormalities because we hypothesize that these will give a direct measure of a patient's capacity to compensate or recover from neurological disease. The second goal is to correlate psychophysical parameters of motor performance and motor learning to the degree of expression of brain networks as revealed by functional imaging. This will provide considerable insight into the brain mechanisms underlying abnormalities in motor control. HD and ITD each have an established genetic basis allowing asymptomatic carriers to be identified. Our preliminary studies indicate that these subjects have psychophysical and network abnormalities even though more conventional assessments fail to find any evidence of neurological disturbance. If we confirm and extend these observations, then we will have the tools to follow therapeutic intervention at the earliest stages of disease. In the long term, we hope that our work will lead to the development of a battery of motor tasks that, in conjunction with functional imaging, will be applicable to any neurological disease. This battery will quantify motor deficits; allow monitoring of therapy; provide insight into brain mechanisms: and may serve a rehabilitative function. Educational Plan: By completing this project, I will accomplish six main educational objectives: (1) The design and application of motor psychophysics to neurological disease; (2) Learning PET imaging techniques; (3) The development of skills in brain network analysis and modeling; (4) Learning advanced statistical methods; (5) Learning the principles of functional magnetic resonance imaging; (5) Learning to conduct ethical scientific research, collaborate with colleagues, and produce high quality presentations and publications; and (6) Learning to place experimental findings within their clinical context, and relate them to clinical assessments.

DESCRIPTION (provided by applicant): Approximately 80% of stroke patients experience acute hemiparesis and approximately 40% have chronic hemiparesis. This impairment significantly limits functional use of the arm. Traditionally, hemiparesis has been characterized by weakness, spasticity and unwanted synergies. These abnormalities are usually assessed using clinical scales. Here we propose to approach the problem of hemiparesis from a more quantitative and modular motor control perspective. Our previous work, investigating planar reaching movements in healthy young subjects, demonstrates that accurate reaching depends en two types of visuomotor transformation. The first transforms a visual target into a planned trajectory in vectorial space centered at the hand with independent specification of direction and extent. This spatial transformation requires the learning of a task-specific reference frame and scaling factor. Once a plan in extrinsic coordinates is determined, a second transformation generates the required torques in intrinsic (joint) coordinates. This dynamic transformation requires learning of the biomechanical properties of the limb. The principal hypothesis of this proposal is that patients with hemiparesis have impairments in using and learning visuomotor transformations. Our secondary hypotheses are that there will be differential impairments in these visuomotor transformations depending on whether the dominant or non-dominant arm is affected and on lesion location. We will include patients who are six months or more out from their first stroke with clinically demonstrable arm weakness at stroke onset. All patients in the study will have had a structural brain MRI with diffusion and perfusion-weighted imaging. Patients with more than one lesion or inability to follow instructions will be excluded. Patients (and age-matched controls) will perform planar reaching movements with their arm supported on a horizontal surface to eliminate gravity. Hand position and joint angle data will be obtained using a magnetic recording system. Motor learning will be assessed by having subjects adapt to a cursor rotation or a laterally-placed mass. Lesion location will be determined by a neuroradiologist. Confirmation of common lesion locations across patients will be made by transforming the T1 images into Tailarach space using SPM2. Our findings will contribute to a mechanistic understanding of hemiparesis, suggest novel rehabilitation strategies and help find better measures of recovery.

The overall goal of this study is to use continuous arterial spin labeling, CASL, perfusion MRI to characterize diaschisis and determine its contribution to hemiparesis after subacute stroke. Stroke remains the leading cause of death worldwide despite improvements in acute stroke treatment and prevention. Patients show considerable variability in the degree to which they recover and the mechanisms underlyingthis variabilityare still largely unknown. One mechanism, termed diaschisis, was proposed almost a century ago to explain part of the initial deficit and the ability topartially recover from it. Diaschisis refers to a reduction of cerebral blood flow in uninjured brain regions that are connected to the area of injury. Animal models suggest that the reduction in cerebral blood flow is the result of decreased neuronal activity in the connected areas. Withinthis framework, the initial deficit is due to combination of diaschisis and the loss of function at the lesion location, whereas recovery occurs as diaschisis reverses. Previous attempts to study this phenomenon in humans have relied on PET and SPECT, which are expensive, not widelyavailable, and require injection of radioactive tracers, limitingthe frequency and practicality of use. As a result, there have been relatively few studies of diaschisis and thus its relevance to stroke deficit and stroke recovery remains to be determined. Here we propose to use arterial spin labeling, ASL, to investigatediaschisis in patients in the subacute and then chronic phases after hemiparetic stroke. ASL is non-invasive MR method that quantifies blood flow. To date it has been used to measure cerebral blood flow, CBF, that has been compromised by vascular occlusive disease. Here we propose a novel application of ASL for detection of decreases in CBF due to reduction in neuronal activity. The main aim of this grant is to demonstrate reductions in CBF in the contralateral cerebellum (crossed cerebellar diaschisis)and in the contralesional hemisphere. We will then correlate these changes with initial motor deficit in 3 to 4 weeks after stroke and with degree of recovery at 6 months.

DESCRIPTION (provided by applicant): It is well known that most movements improve with practice. However, the way the brain accomplishes this remains largely unknown. Functional magnetic resonance imaging (fMRI) has provided significant insights into brain mechanisms of cognition. In comparison, fMRI studies of motor learning and motor memory have been limited by the constrained space and the problem of unwanted head movements. As a result the great majority of fMRI studies of motor learning have investigated sequences of finger movements. To overcome these limitations, we have developed a novel MR-compatible wrist task to study two types of motor learning that undrlie our ability to accurately point to visual targets: (1) Motor skill learning - the ability to acquire new patterns of muscle activity so that movement accuracy increases without a reduction in speed. (2) Visuomotor adaptation - the ability to associate an already learned pattern of muscle activation with a new spatial goal. We will study these two types of learning with a baseline and a rotation condition, respectively. In the baseline condition, subjects make fast uncorrected pointing movements of the wrist to a series of 8 targets arrayed on a screen. Skill is the decrease in movement time (MT) and endpoint variability with practice. In the rotation condition, with the same target set, subjects initially make systematic 30[unreadable] directional errors, which they reduce through adaptation. Importantly, we have developed a method that keeps MT and peak velocity constant so that any observed brain activation changes are attributable to learning and not performance changes. In addition, we will also examine the neural correlates of re-learning on a second day, differences between the dominant and non-dominant hand, mechanisms of motor interference, and degree of transfer of motor learning between arms. A better understanding of hemispheric contributions to motor learning and motor memory will offer insight into mechanisms of recovery after focal brain injury. We anticipate our approach will be applicable to patients in future studies.

Whether stepping on the brake pedal when a hazard appears in the road ahead, beginning a sprint in response to the starter's pistol, or reaching to catch a tipping glass, we possess a remarkable ability to respond rapidly and accurately to demands for action. These responses occur within a reaction time of around 250 milliseconds, during which we must perceive the need to act, decide precisely what action to take, and generate motor commands to coordinate the activity of dozens of muscles to make the movement. It is currently not clearly understood what steps in this pathway take the greatest amount of time, which brain regions are responsible for which computations, and whether or not the reaction time could be reduced even lower. With support from the National Science Foundation, Dr. John Krakauer and colleagues Dr. Adrian Haith, Dr. Pablo Celnik and Dr. Joshua Ewan will use innovative behavioral experiments with human participants to stress the limits of the reaction time and tease apart its constituent processes and limitations. They will record brain activity during these same tasks (using EEG) in order to identify the underlying brain networks responsible and will apply non-invasive brain stimulation to establish the specific function of these brain networks in generating rapid motor responses.

A number of neurological disorders, such as Parkinson disease, are associated with slowing of motor responses. Such symptoms are not well-understood at present. The knowledge gained from this study will yield important insights as to why this slowing may occur and what, if anything, could be done to treat it from a therapeutic and/or clinical perspective. A deeper understanding of the reaction time will also be of significant interest to those seeking to push the limits of human performance, such as athletes and race car drivers, or amateurs seeking to better their own performance. Furthermore, data collected in this study are likely to be of broad interest and therefore will be made available to other researchers. EEG recordings will be of particular interest to researchers developing advanced analytical techniques for mapping human brain activity. Finally, students from under-represented minorities, including those from economically disadvantaged backgrounds, will have the opportunity to directly participate in this project and gain valuable research experience through the Johns Hopkins Diversity Summer Internship Program.

A stroke often damages motor areas of the brain. Understandably, this leads to a loss of movement control: the limbs become weak, and movements are slower and less well-coordinated. In addition to loss of function, patients also gain unwanted muscle contractions called synergies. For example, whenever the arm is lifted (shoulder abduction), the elbow flexes. These co-contractions intrude into normal movements. Synergies, not just weakness or lack of control, are a major contributor to disability in stroke survivors. Many previous studies have investigated stroke recovery in animals (typically monkeys because of the close similarities of their motor system to humans), but these have focused on recovery of lost function, not on synergies. One reason is that in most previous work monkeys did not express overt synergies; until now we have therefore lacked a model of one of the major causes of post-stroke disability. This critical gap in our understanding has largely gone unnoticed. We need to know how to induce synergies in monkeys, which neural circuits are responsible for them, how they are controlled in health, and how this control becomes disordered after stroke. This project seeks to address this gap, paving the way for a rational approach to new therapy for synergies. In the first experiment, monkeys will be trained on a reaching task, and then implanted with electrodes to measure muscle activity. High speed video recordings will extract movement kinematics. An instrumented linear motor will measure tendon-tap reflexes. After baseline recordings, we will induce a focal cortical ischemic lesion, and gather further data over the subsequent months. We will measure the development of inappropriate contractions of elbow flexors with shoulder abductors during outward reaches. We will analyze reaching trajectories to quantify quality of movement (equivalent to a dexterity measure in the hand, but for reach). Tendon tap reflexes will assess spasticity. Lesions of five different cortical regions will be compared. The lesion which produces the most severe synergy will then be combined with damage to the magnocellular red nucleus, which we hypothesize will further accentuate synergy expression. This experiment will elucidate the detailed functional anatomy of the post-stroke syndrome, and also yield an optimized monkey model of pathological synergies. In the second experiment, monkeys will be trained to move an on-screen cursor controlled by shoulder abduction-elbow flexion torques into targets, allowing parametric examination of independent versus co- activation. Initially neural circuits will be characterized in healthy monkeys. After necessary surgical implants, neural activity will be recorded from different parts of the motor cortex, the reticular formation, and the spinal cord. We hypothesize that spinal circuits will show neural activity consistent with co-activation of shoulder and elbow muscles to generate synergies; activity in supraspinal areas will be consistent with either driving this spinal circuit, or suppressing it to allow independent muscle activation. Recordings will then be repeated in monkeys subjected to the lesion which generates optimal synergies, to reveal the nature of pathological changes.

Much of our daily behavior is supported by habits, which we usually only become aware of when our behavior needs to change. For example, if we need to switch to driving a right-hand-drive vehicle while on vacation, we find ourselves habitually looking in the wrong location for the rear-view mirror and reaching with the wrong arm for the handbrake. Habits can be beneficial because they enable skills to be performed well without much conscious thought. However, habits can also be detrimental because they limit our capacity to change our behavior. It is therefore critical to understand how we acquire and eliminate habits. Habit formation has been widely studied using simple tasks in which people learn to associate visual cues with specific responses. This body of research has given rise to an “all or nothing” view of habits, in which behavior is either flexible or habitual. In more complex skills, however, some aspects of our behavior might be habitual while others remain flexible. For example, a worker in a production facility might have substantial flexibility in being able to assemble many different types of devices. But if the layout of their workspace is re-organized, they may habitually reach for the old (now wrong) location to retrieve a needed tool or component. It can therefore be critical to ensure that seemingly flexible behavior does not mask latent habits. The goal of this project is to develop systematic approaches to identifying which aspects of a behavior are habitual, understand how and when different components of a complex skill become habitual with practice and determine how formation of different types of habits can be shaped based on how we practice. Being able to identify and target specific habits during learning will help improve training programs to foster good habits and avoid or eliminate bad ones. While this project will focus on the role of habits in learning complex skills, the findings will also be applicable to other domains where habits play an important role, such as understanding consumer behavior, promoting a healthy lifestyle, and developing strategies for rehabilitation from injury or disease.<br/><br/>In this project, human participants will learn and practice a variety of tasks which require them to use multiple component processes to determine appropriate responses to visual stimuli. Over the course of practice over multiple sessions, the investigators will track whether and when the different underlying components of behavior become habitual. Whether or not each component of behavior becomes habitual is assessed by altering the requirements of the task in a way that is specific to each component, and then measuring the incidence of habitual ‘slips-of-action’ in which participants revert to the pattern of behavior they originally practiced. In further experiments, participants will be trained under conditions designed to promote formation of one type of habit while inhibiting formation of others.<br/><br/>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.