Optimal reaching subject to computational and physical constraints reveals structure of the sensorimotor control system.

Optimal feedback control provides an abstract framework describing the architecture of the sensorimotor system without prescribing implementation details such as what coordinate system to use, how feedback is incorporated, or how to accommodate changing task complexity. We investigate how such details are determined by computational and physical constraints by creating a model of the upper limb sensorimotor system in which all connection weights between neurons, feedback, and muscles are unknown. By optimizing these parameters with respect to an objective function, we find that the model exhibits a preference for an intrinsic (joint angle) coordinate representation of inputs and feedback and learns to calculate a weighted feedforward and feedback error. We further show that complex reaches around obstacles can be achieved by augmenting our model with a path-planner based on via points. The path-planner revealed "avoidance" neurons that encode directions to reach around obstacles and "placement" neurons that make fine-tuned adjustments to via point placement. Our results demonstrate the surprising capability of computationally constrained systems and highlight interesting characteristics of the sensorimotor system.

Automatic learning mechanisms for flexible human locomotion.

Movement flexibility and automaticity are necessary to successfully navigate different environments. When encountering difficult terrains such as a muddy trail, we can change how we step almost immediately so that we can continue walking. This flexibility comes at a cost since we initially must pay deliberate attention to how we are moving. Gradually, after a few minutes on the trail, stepping becomes automatic so that we do not need to think about our movements. Canonical theory indicates that different adaptive motor learning mechanisms confer these essential properties to movement: explicit control confers flexibility, while forward model recalibration confers automaticity. Here we uncover a distinct mechanism of treadmill walking adaptation - an automatic stimulus-response mapping - that confers both properties to movement. The mechanism is flexible as it learns stepping patterns that can be rapidly changed to suit a range of treadmill configurations. It is also automatic as it can operate without deliberate control or explicit awareness by the participants. Our findings reveal a tandem architecture of forward model recalibration and automatic stimulus-response mapping mechanisms for walking, reconciling different findings of motor adaptation and perceptual realignment.

Mapping subcortical motor pathways in humans with startle-conditioned TMS.

Subcortical motor pathways, such as the reticulospinal tract, are critical for producing and modulating voluntary movements and have been implicated in neurological conditions. Previous research has described the presence of ipsilateral motor evoked potentials (iMEPs) in the arm to transcranial magentic stimulation (TMS), and suggested they could be mediated by the uncrossed corticospinal tract or by ipsilateral cortico-reticulospinal connections. Here, we sought to elucidate the role of the reticulospinal tract in mediating iMEPs by assessing their modulation by a startling acoustic stimulus and mapping these responses across multiple upper limb effectors. In a first experiment, we delivered TMS at various intervals (1, 5, 10 and 15 ms) after a startling acoustic stimulus, known to excite the reticular formation, to elicit iMEPs in the arm. We observed robust facilitation of iMEP area when startle conditioning preceded TMS at the 10 ms interval. In a second experiment, we replicated our findings showing that both the area and number of iMEPs in the arm increases with startle conditioning. Using this technique, we observed that iMEPs are more prominent in the arm compared with the hand. In a third experiment, we also observed greater presence of iMEPs in flexor compared with extensor muscles. Together, these findings are consistent with properties of the reticulospinal tract observed in animals, suggesting that iMEPs primarily reflect reticulospinal activity. Our findings imply that we can use this approach to track modulation of cortico-reticulospinal excitability following interventions or neurological conditions where the reticulospinal tract may be involved in motor recovery.

A novel video game for remote studies of motor adaptation in children.

Here we designed a motor adaptation video game that could be played remotely (at home) through a web browser. This required the child to adapt to a visuomotor rotation between their hand movement and a ball displayed in the game. The task had several novel features, specifically designed to allow the study of the developmental trajectory of adaptation across a wide range of ages. We test the concurrent validity by comparing children's performance on our remote task to the same task performed in the laboratory. All participants remained engaged and completed the task. We quantified feedforward and feedback control during this task. Feedforward control, a key measure of adaptation, was similar at home and in the laboratory. All children could successfully use feedback control to guide the ball to a target. Traditionally, motor learning studies are performed in a laboratory to obtain high quality kinematic data. However, here we demonstrate concurrent validity of kinematic behavior when conducted at home. Our online platform provides the flexibility and ease of collecting data that will enable future studies with large sample sizes, longitudinal experiments, and the study of children with rare diseases.

Joint-level coordination patterns for split-belt walking across different speed ratios.

Locomotion is a highly flexible process, requiring rapid changes to gait due to changes in the environment or goals. Here, we used a split-belt treadmill to examine how the central nervous system coordinates a novel gait pattern. Existing research has focused on summary measures, most often step lengths, when describing changes induced while walking on the split-belt treadmill and during subsequent aftereffects. Here, we asked how the nervous system adjusts individual joint motions and the coordination pattern of the legs when people walk with one leg moving at either 2×, 3×, or 4× the speed of the other leg. We found that relative to tied-belt walking, split-belt perturbations change the timing relationships between the legs while most joint angle peaks and range of motion change little. The kinematic changes over the course of adaptation (i.e., from the beginning to end of a single split-belt walking bout) were subtle, particularly when comparing individual joint motions. The magnitude of the belt speed differences impacted intralimb coordination but did not produce consistent differences in most other measures. Most significant changes in kinematics occurred in the fast leg. Overall, interlimb timing changes drove a large proportion of the differences observed between tied-belt and split-belt gaits. Thus, it appears that the central nervous system can produce novel gait patterns through changes in coordination between legs that lead to new configurations at significant time points. These patterns can use within-limb and within-joint patterns that closely resemble those of normal walking.NEW & NOTEWORTHY We studied how the nervous system coordinates limb movements during asymmetric gait. Using a split-belt treadmill, we found that most changes in motion occurred when comparing motions between limbs, rather than among joints within a limb. Individual joint patterns resembled speed-matched comparisons, but this meant that joint movements became asymmetric during split-belt walking. These findings demonstrate that the nervous system can use consistent joint motions that are reconfigured in time to achieve new gait patterns.

Reinforcement Learning Is Impaired in the Sub-acute Post-stroke Period.

Background: Neurorehabilitation approaches are frequently predicated on motor learning principles. However, much is left to be understood of how different kinds of motor learning are affected by stroke causing hemiparesis. Here we asked if two kinds of motor learning often employed in rehabilitation, (1) reinforcement learning and (2) error-based adaptation, are altered at different times after stroke. Methods: In a cross-sectional design, we compared learning in two groups of patients with stroke, matched for their baseline motor execution deficit on the paretic side. The early group was tested within 3 months following stroke (N = 35) and the late group was tested more than 6 months after stroke (N = 30). Two types of task were studied: one based on reinforcement learning and the other on error-based learning. Results: We found that reinforcement learning was impaired in the early but not the late group, whereas error-based learning was unaffected compared to controls. These findings could not be attributed to differences in baseline execution, cognitive impairment, gender, age, or lesion volume and location. Conclusions: The presence of a specific impairment in reinforcement learning in the first 3 months after stroke has important implications for rehabilitation. It might be necessary to either increase the amount of reinforcement feedback given early or even delay onset of certain forms of rehabilitation training, e.g., like constraint-induced movement therapy, and instead emphasize others forms of motor learning in this early time period. A deeper understanding of stroke-related changes in motor learning capacity has the potential to facilitate the development of new, more precise treatment interventions.

Motor control: In constant pursuit of optimality.

Humans learn through exploration. A new study suggests that this may be how we learn to save energy when we walk.

Context-Specificity of Locomotor Learning Is Developed during Childhood.

Humans can perform complex movements with speed and agility in the face of constantly changing task demands. To accomplish this, motor plans are adapted to account for errors in our movements because of changes in our body (e.g., growth or injury) or in the environment (e.g., walking on sand vs ice). It has been suggested that adaptation that occurs in response to changes in the state of our body will generalize across different movement contexts and environments, whereas adaptation that occurs with alterations in the external environment will be context-specific. Here, we asked whether the ability to form generalizable versus context-specific motor memories develops during childhood. We performed a cross-sectional study of context-specific locomotor adaptation in 35 children (3-18 years old) and 7 adults (19-31 years old). Subjects first adapted their gait and learned a new walking pattern on a split-belt treadmill, which has two belts that move each leg at a different speed. Then, subjects walked overground to assess the generalization of the adapted walking pattern across different environments. Our results show that the generalization of treadmill after-effects to overground walking decreases as subjects' age increases, indicating that age and experience are critical factors regulating the specificity of motor learning. Our results suggest that although basic locomotor patterns are established by two years of age, brain networks required for context-specific locomotor learning are still being developed throughout youth.

Dissociation between abnormal motor synergies and impaired reaching dexterity after stroke.

Most patients with stroke experience motor deficits, usually referred to collectively as hemiparesis. Although hemiparesis is one of the most common and clinically recognizable motor abnormalities, it remains undercharacterized in terms of its behavioral subcomponents and their interactions. Hemiparesis comprises both negative and positive motor signs. Negative signs consist of weakness and loss of motor control (dexterity), whereas positive signs consist of spasticity, abnormal resting posture, and intrusive movement synergies (abnormal muscle co-activations during voluntary movement). How positive and negative signs interact, and whether a common mechanism generates them, remains poorly understood. Here, we used a planar, arm-supported reaching task to assess poststroke arm dexterity loss, which we compared with the Fugl-Meyer stroke scale; a measure primarily reflecting abnormal synergies. We examined 53 patients with hemiparesis after a first-time ischemic stroke. Reaching kinematics were markedly more impaired in patients with subacute (<3 mo) compared to chronic (>6 mo) stroke even for similar Fugl-Meyer scores. This suggests a dissociation between abnormal synergies (reflected in the Fugl-Meyer scale) and loss of dexterity, which in turn suggests different underlying mechanisms. Moreover, dynamometry suggested that Fugl-Meyer scores capture weakness as well as abnormal synergies, in line with these two deficits sharing a neural substrate. These findings have two important implications: First, clinical studies that test for efficacy of rehabilitation interventions should specify which component of hemiparesis they are targeting and how they propose to measure it. Metrics used widely for this purpose may not always be chosen appropriately. For example, as we show here, the Fugl-Meyer score may capture some hemiparesis components (abnormal synergies and weakness) but not others (loss of dexterity). Second, there may be an opportunity to design rehabilitation interventions to address specific subcomponents of hemiparesis.NEW & NOTEWORTHY Motor impairment is common after stroke and comprises reduced dexterity, weakness, and abnormal muscle synergies. Here we report that, when matched on an established synergy and weakness scale (Fugl-Meyer), patients with subacute stroke have worse reaching dexterity than chronic ones. This result suggests that the components of hemiparesis are dissociable and have separable mechanisms and, thus, may require distinct assessments and treatments.

Younger and Late Middle-Aged Adults Exhibit Different Patterns of Cognitive-Motor Interference During Locomotor Adaptation, With No Disruption of Savings.

It has been proposed that motor adaptation and subsequent savings (or faster relearning) of an adapted movement pattern are mediated by cognitive processes. Here, we evaluated the pattern of cognitive-motor interference that emerges when young and late middle-aged adults perform an executive working memory task during locomotor adaptation. We also asked if this interferes with savings of a newly learned walking pattern, as has been suggested by a study of reaching adaptation. We studied split-belt treadmill adaptation and savings in young (21 ± 2 y/o) and late middle-aged (56 ± 6 y/o) adults with or without a secondary 2-back task during adaptation. We found that young adults showed similar performance on the 2-back task during baseline and adaptation, suggesting no effect of the dual-task on cognitive performance; however, dual-tasking interfered with adaptation over the first few steps. Conversely, dual-tasking caused a decrement in cognitive performance in late middle-aged adults with no effect on adaptation. To determine if this effect was specific to adaptation, we also evaluated dual-task interference in late middle-aged adults that dual-tasked while walking in a complex environment that did not induce motor adaptation. This group exhibited less cognitive-motor interference than late middle-aged adults who dual-tasked during adaptation. Savings was unaffected by dual-tasking in both young and late middle-aged adults, which may indicate different underlying mechanisms for savings of reaching and walking. Collectively, our findings reveal an age-dependent effect of cognitive-motor interference during dual-task locomotor adaptation and no effect of dual-tasking on savings, regardless of age. Young adults maintain cognitive performance and show a mild decrement in locomotor adaptation, while late middle-aged adults adapt locomotion at the expense of cognitive performance.

Training at asymptote stabilizes motor memories by reducing intracortical excitation.

Learning similar motor skills in close succession is limited by interference, a phenomenon that takes place early after acquisition when motor memories are unstable. Interference can be bidirectional, as the first memory can be disrupted by the second (retrograde interference), or the second memory can be disrupted by the first (anterograde interference). The heightened plastic state of primary motor cortex after learning is thought to underlie interference, as unstable motor memories compete for neural resources. While time-dependent consolidation processes reduce interference, the passage of time (~6 h) required for memory stabilization limits our capacity to learn multiple motor skills at once. Here, we demonstrate in humans that prolonged training at asymptote of an initial motor skill reduces both retrograde and anterograde interference when a second motor skill is acquired in close succession. Neurophysiological assessments via transcranial magnetic stimulation reflect this online stabilization process. Specifically, excitatory neurotransmission in primary motor cortex increased after short training and decreased after prolonged training at performance asymptote. Of note, this reduction in intracortical excitation after prolonged training was proportional to better skill retention the following day. Importantly, these neurophysiological effects were not observed after motor practice without learning or after a temporal delay. Together, these findings indicate that prolonged training at asymptote improves the capacity to learn multiple motor skills in close succession, and that downregulation of excitatory neurotransmission in primary motor cortex may be a marker of online motor memory stabilization.

Reinforcement Signaling Can Be Used to Reduce Elements of Cerebellar Reaching Ataxia.

Damage to the cerebellum causes a disabling movement disorder called ataxia, which is characterized by poorly coordinated movement. Arm ataxia causes dysmetria (over- or under-shooting of targets) with many corrective movements. As a result, people with cerebellar damage exhibit reaching movements with highly irregular and prolonged movement paths. Cerebellar patients are also impaired in error-based motor learning, which may impede rehabilitation interventions. However, we have recently shown that cerebellar patients can learn a simple reaching task using a binary reinforcement paradigm, in which feedback is based on participants' mean performance. Here, we present a pilot study that examined whether patients with cerebellar damage can use this reinforcement training to learn a more complex motor task-to decrease the path length of their reaches. We compared binary reinforcement training to a control condition of massed practice without reinforcement feedback. In both conditions, participants made target-directed reaches in 3-dimensional space while vision of their movement was occluded. In the reinforcement training condition, reaches with a path length below participants' mean were reinforced with an auditory stimulus at reach endpoint. We found that patients were able to use reinforcement signaling to significantly reduce their reach paths. Massed practice produced no systematic change in patients' reach performance. Overall, our results suggest that binary reinforcement training can improve reaching movements in patients with cerebellar damage and the benefit cannot be attributed solely to repetition or reduced visual control.

Cerebellar patients have intact feedback control that can be leveraged to improve reaching.

It is thought that the brain does not simply react to sensory feedback, but rather uses an internal model of the body to predict the consequences of motor commands before sensory feedback arrives. Time-delayed sensory feedback can then be used to correct for the unexpected-perturbations, motor noise, or a moving target. The cerebellum has been implicated in this predictive control process. Here, we show that the feedback gain in patients with cerebellar ataxia matches that of healthy subjects, but that patients exhibit substantially more phase lag. This difference is captured by a computational model incorporating a Smith predictor in healthy subjects that is missing in patients, supporting the predictive role of the cerebellum in feedback control. Lastly, we improve cerebellar patients' movement control by altering (phase advancing) the visual feedback they receive from their own self movement in a simplified virtual reality setup.

Individualized feedback to change multiple gait deficits in chronic stroke.

BACKGROUND: Walking deficits in people post-stroke are often multiple and idiosyncratic in nature. Limited patient and therapist resources necessitate prioritization of deficits such that some may be left unaddressed. More efficient delivery of therapy may alleviate this challenge. Here, we look to determine the utility of a novel principal component-based visual feedback system that targets multiple, patient-specific features of gait in people post-stroke. METHODS: Ten individuals with stroke received two sessions of visual feedback to attain a walking goal. This goal consisted of bilateral knee and hip joint angles of a typical 'healthy' walking pattern. The feedback system uses principal component analysis (PCA) to algorithmically weight each of the input features so that participants received one stream of performance feedback. In the first session, participants had to explore different patterns to achieve the goal, and in the second session they were informed of the goal walking pattern. Ten healthy, age-matched individuals received the same paradigm, but with a hemiparetic goal (i.e. to produce the pattern of an exemplar stroke participant). This was to distinguish the extent to which performance limitations in stroke were due neurological injury or the PCA based visual feedback itself. RESULTS: Principal component-based visual feedback can differentially bias multiple features of walking toward a prescribed goal. On average, individuals with stroke typically improved performance via increased paretic knee and hip flexion, and did not perform better with explicit instruction. In contrast, healthy people performed better (i.e. could produce the desired exemplar stroke pattern) in both sessions, and were best with explicit instruction. Importantly, the feedback for stroke participants accommodated a heterogeneous set of walking deficits by individually weighting each feature based on baseline walking. CONCLUSIONS: People with and without stroke are able to use this novel visual feedback to train multiple, specific features of gait. Important for stroke, the PCA feedback allowed for targeting of patient-specific deficits. This feedback is flexible to any feature of walking in any plane of movement, thus providing a potential tool for therapists to simultaneously target multiple aberrant features of gait.

Providing low-dimensional feedback of a high-dimensional movement allows for improved performance of a skilled walking task.

Learning a skilled movement often requires changing multiple dimensions of movement in a coordinated manner. Serial training is one common approach to learning a new movement pattern, where each feature is learned in isolation from the others. Once one feature is learned, we move on to the next. However, when learning a complex movement pattern, serial training is not only laborious but can also be ineffective. Often, movement features are linked such that they cannot simply be added together as we progress through training. Thus, the ability to learn multiple features in parallel could make training faster and more effective. When using visual feedback as the tool for changing movement, however, such parallel training may increase the attentional load of training and impair performance. Here, we developed a novel visual feedback system that uses principal component analysis to weight four features of movement to create a simple one-dimensional 'summary' of performance. We used this feedback to teach healthy, young participants a modified walking pattern and compared their performance to those who received four concurrent streams of visual information to learn the same goal walking pattern. We demonstrated that those who used the principal component-based visual feedback improved their performance faster and to a greater extent compared to those who received concurrent feedback of all features. These results suggest that our novel principal component-based visual feedback provides a method for altering multiple features of movement toward a prescribed goal in an intuitive, low-dimensional manner.

Trading Symmetry for Energy Cost During Walking in Healthy Adults and Persons Poststroke.

Background. Humans typically walk in ways that minimize energy cost. Recent work has found that healthy adults will even adopt new ways of walking when a new pattern costs less energy. This suggests potential for rehabilitation to drive changes in walking by altering the energy costs of walking patterns so that the desired pattern becomes energetically optimal (ie, costs least energy of all available patterns). Objective. We aimed to change gait symmetry in healthy adults and persons poststroke by creating environments where changing symmetry allowed the participants to save energy. Methods. Across 3 experiments, we tested healthy adults (n = 12 in experiment 1, n = 20 in experiment 2) and persons poststroke (n = 7 in experiment 3) in a novel treadmill environment that linked asymmetric stepping and gait speed-2 factors that influence energy cost-to create situations where walking with one's preferred gait symmetry (or asymmetry, in the case of the persons poststroke) was no longer the least energetically costly way to walk. Results. Across the 3 experiments, we found that most participants changed their gait when experiencing the new energy landscape. Healthy adults often adopted an asymmetric gait if it saved energy, and persons poststroke often began to step more symmetrically than they prefer to walk in daily life. Conclusions. We used a novel treadmill environment to show that people with and without stroke change clinically relevant features of walking to save energy. These findings suggest that rehabilitation approaches aimed at making symmetric walking energetically "easier" may promote gait symmetry after stroke.

The capacity to learn new motor and perceptual calibrations develops concurrently in childhood.

Learning new movements through an error-based process called motor adaptation is thought to involve multiple mechanisms which are still largely not understood. Previous studies have shown that young children adapt movement more slowly than adults, perhaps supporting the involvement of distinct neural circuits that come online at different stages of development. Recent studies in adults have shown that in addition to recalibrating a movement, motor adaptation also leads to changes in the perception of that movement. However, we do not yet understand the relationship between the processes that underlie motor and perceptual recalibration. Here we studied motor and perceptual recalibration with split-belt walking adaptation in adults and children aged 6-8 years. Consistent with previous work, we found that this group of children adapted their walking patterns more slowly than adults, though individual children ranged from slow to adult-like in their adaptation rates. Perceptual recalibration was also reduced in the same group of children compared to adults, with individual children ranging from having no recalibration to having adult-like recalibration. In sum, faster motor adaptation and the ability to recalibrate movement perception both come online within a similar age-range, raising the possibility that the same sensorimotor mechanisms underlie these processes.

The cerebellum as a movement sensor.

In this article, we review a broad range of studies of cerebellar function and dysfunction and interpret them within the framework that the cerebellum acts as part of a mechanism of predictive control. We describe studies that span human behaviour and consider the motor and sensory impairments that result from cerebellar damage. We conclude that a parsimonious explanation of cerebellar function is as a predictor of the sensory outcomes of movement. However, future studies are needed to more rigorously test this hypothesis and determine how the cerebellar circuit might perform this type of computation.

Patients with Cerebellar Ataxia Do Not Benefit from Limb Weights.

Patients with cerebellar ataxia are sometimes treated by the addition of mass to the limbs, though this practice has received limited study. Recent work suggests that adding mass to the limbs might have predictable effects on the pattern of cerebellar dysmetria (i.e., over or undershooting) that depends on a hypothesized mismatch between the actual limb inertia and the brain's estimate of limb inertia. Based on this model, we predicted that addition of mass would only be effective in reducing dysmetria in hypometric patients. Cerebellar patients were challenged with making a single-joint, single degree of freedom reaching movement while various limb masses were tested. In this task, some single-jointed reaches were improved by adding masses that were optimized in a patient-specific manner. However, this improvement did not translate to multi-joint movements. In multi-joint movements, the "best" patient-specific masses (as determined in a single-joint task) generally exacerbated subjects' reaching errors. This finding raises questions as to the merits of adding limb weights as a therapy to mitigate the effects of dysmetria.

Movement and perception recalibrate differently across multiple days of locomotor learning.

Learning a new movement through error-based adaptation leads to recalibration of movement and altered perception of that movement. Although presumed to be closely related, the relationship between adaptation-based motor and perceptual changes is not well understood. Here we investigated the changes in motor behavior and leg speed perception over 5 days of split-belt treadmill adaptation. We specifically wanted to know if changes in the perceptual domain would demonstrate savings-like behavior (i.e., less recalibration with more practice) and if these changes would parallel the savings observed in the motor domain. We found that the recalibration of leg speed perception decreased across days of training, indicating savings-like behavior in this domain. However, we observed that the magnitude of savings across days was different between motor and perceptual domains. These findings suggest a degree of independence between the motor and perceptual processes that occur with locomotor adaptation. NEW & NOTEWORTHY Error-based adaptation learning drives changes in movement and perception of movement. Are these changes across domains linked or simply coincidental? Here, we studied changes in movement and perception across 5 days of repeated locomotor adaptation. Savings-like behavior in the motor and perceptual domains developed with different magnitudes and over different timescales, leading us to conclude that motor and perceptual processes operate at least somewhat independently during locomotor adaptation.