Implicit adaptation compensates for erratic explicit strategy in human motor learning

Perceptual error based on Bayesian cue combination drives implicit motor adaptation

The sensorimotor system can recalibrate itself without our conscious awareness, a type of procedural learning whose computational mechanism remains undefined. Recent findings on implicit motor adaptation, such as over-learning from small perturbations and fast saturation for increasing perturbation size, challenge existing theories based on sensory errors. We argue that perceptual error, arising from the optimal combination of movement-related cues, is the primary driver of implicit adaptation. Central to our theory is the increasing sensory uncertainty of visual cues with increasing perturbations, which was validated through perceptual psychophysics (Experiment 1). Our theory predicts the learning dynamics of implicit adaptation across a spectrum of perturbation sizes on a trial-by-trial basis (Experiment 2). It explains proprioception changes and their relation to visual perturbation (Experiment 3). By modulating visual uncertainty in perturbation, we induced unique adaptation responses in line with our model predictions (Experiment 4). Overall, our perceptual error framework outperforms existing models based on sensory errors, suggesting that perceptual error in locating one's effector, supported by Bayesian cue integration, underpins the sensorimotor system's implicit adaptation.

Reward actively engages both implicit and explicit components in dual force field adaptation

Motor learning occurs through multiple mechanisms, including unsupervised, supervised (error based), and reinforcement (reward based) learning. Although studies have shown that reward leads to an overall better motor adaptation, the specific processes by which reward influences adaptation are still unclear. Here, we examine how the presence of reward affects dual adaptation to novel dynamics and distinguish its influence on implicit and explicit learning. Participants adapted to two opposing force fields in an adaptation/deadaptation/error-clamp paradigm, where five levels of reward (a score and a digital face) were provided as participants reduced their lateral error. Both reward and control (no reward provided) groups simultaneously adapted to both opposing force fields, exhibiting a similar final level of adaptation, which was primarily implicit. Triple-rate models fit to the adaptation process found higher learning rates in the fast and slow processes and a slightly increased fast retention rate for the reward group. Whereas differences in the slow learning rate were only driven by implicit learning, the large difference in the fast learning rate was mainly explicit. Overall, we confirm previous work showing that reward increases learning rates, extending this to dual-adaptation experiments and demonstrating that reward influences both implicit and explicit adaptation. Specifically, we show that reward acts primarily explicitly on the fast learning rate and implicitly on the slow learning rates.NEW & NOTEWORTHY Here we show that rewarding participants' performance during dual force field adaptation primarily affects the initial rate of learning and the early timescales of adaptation, with little effect on the final adaptation level. However, reward affects both explicit and implicit components of adaptation. Whereas the learning rate of the slow process is increased implicitly, the fast learning and retention rates are increased through both implicit components and the use of explicit strategies.

Imprecise perception of hand position during early motor adaptation

Localizing one's body parts is important for movement control and motor learning. Recent studies have shown that the precision with which people localize their hand places constraints on motor adaptation. Although these studies have assumed that hand localization remains equally precise across learning, we show that precision decreases rapidly during early motor learning. In three experiments, healthy young participants (n = 92) repeatedly adapted to a 45° visuomotor rotation for a cycle of two to four reaches, followed by a cycle of two to four reaches with veridical feedback. Participants either used an aiming strategy that fully compensated for the rotation (experiment 1), or always aimed directly at the target, so that adaptation was implicit (experiment 2). We omitted visual feedback for the last reach of each cycle, after which participants localized their unseen hand. We observed an increase in the variability of angular localization errors when subjects used a strategy to counter the visuomotor rotation (experiment 1). This decrease in precision was less pronounced in the absence of reaiming (experiment 2), and when subjects knew that they would have to localize their hand on the upcoming trial, and could thus focus on hand position (experiment 3). We propose that strategic reaiming decreases the precision of perceived hand position, possibly due to attention to vision rather than proprioception. We discuss how these dynamics in precision during early motor learning could impact on motor control and shape the interplay between implicit and strategy-based motor adaptation.NEW & NOTEWORTHY Recent studies indicate that the precision with which people localize their hand limits implicit visuomotor learning. We found that localization precision is not static, but decreases early during learning. This decrease is pronounced when people apply a reaiming strategy to compensate for a visuomotor perturbation and is partly resistant to allocation of attention to the hand. We propose that these dynamics in position sense during learning may influence how implicit and strategy-based motor adaption interact.

Prediction error in implicit adaptation during visually- and memory-guided reaching tasks

Human movements are adjusted by motor adaptation in order to maintain their accuracy. There are two systems in motor adaptation, referred to as explicit or implicit adaptation. It has been suggested that the implicit adaptation is based on the prediction error and has been used in a number of motor adaptation studies. This study aimed to examine the effect of visual memory on prediction error in implicit visuomotor adaptation by comparing visually- and memory-guided reaching tasks. The visually-guided task is thought to be implicit learning based on prediction error, whereas the memory-guided task requires more cognitive processes. We observed the adaptation to visuomotor rotation feedback that is gradually rotated. We found that the adaptation and retention rates were higher in the visually-guided task than in the memory-guided task. Furthermore, the delta-band power obtained by electroencephalography (EEG) in the visually-guided task was increased immediately following the visual feedback, which indicates that the prediction error was larger in the visually-guided task. Our results show that the visuomotor adaptation is enhanced in the visually-guided task because the prediction error, which contributes update of the internal model, was more reliable than in the memory-guided task. Therefore, we suggest that the processing of the prediction error is affected by the task-type, which in turn affects the rate of the visuomotor adaptation.

The cerebellum acts as the analog to the medial temporal lobe for sensorimotor memory

The cerebellum is critical for sensorimotor learning. The specific contribution that it makes, however, remains unclear. Inspired by the classic finding that, for declarative memories, medial temporal lobe structures provide a gateway to the formation of long-term memory but are not required for short-term memory, we hypothesized that, for sensorimotor memories, the cerebellum may play an analogous role. Here we studied the sensorimotor learning of individuals with severe ataxia from cerebellar degeneration. We dissected the memories they formed during sensorimotor learning into a short-term temporally-volatile component, that decays rapidly with a time constant of just 15-20sec and thus cannot lead to long-term retention, and a longer-term temporally-persistent component that is stable for 60 sec or more and leads to long-term retention. Remarkably, we find that these individuals display dramatically reduced levels of temporally-persistent sensorimotor memory, despite spared and even elevated levels of temporally-volatile sensorimotor memory. In particular, we find both impairment that systematically increases with memory window duration over shorter memory windows (<12 sec) and near-complete impairment of memory maintenance over longer memory windows (>25 sec). This dissociation uncovers a new role for the cerebellum as a gateway for the formation of long-term but not short-term sensorimotor memories, mirroring the role of the medial temporal lobe for declarative memories. It thus reveals the existence of distinct neural substrates for short-term and long-term sensorimotor memory, and it explains both newly-identified trial-to-trial differences and long-standing study-to-study differences in the effects of cerebellar damage on sensorimotor learning ability.

Significance Statement

A key discovery about the neural underpinnings of memory, made more than half a century ago, is that long-term, but not short-term, memory formation depends on neural structures in the brain's medial temporal lobe (MTL). However, this dichotomy holds only for declarative memories - memories for explicit facts such as names and dates - as long-term procedural memories - memories for implicit knowledge such as sensorimotor skills - are largely unaffected even with substantial MTL damage. Here we demonstrate that the formation of long-term, but not short-term, sensorimotor memory depends on a neural structure known as the cerebellum, and we show that this finding explains the variability previously reported in the extent to which cerebellar damage affects sensorimotor learning.

Subtle Visual Latency Can Profoundly Impair Implicit Sensorimotor Learning

Short sub-100ms visual feedback latencies are common in many types of human-computer interactions yet are known to markedly reduce performance in a wide variety of motor tasks from simple pointing to operating surgical robotics. These latencies are also present in the computer-based experiments used to study the sensorimotor learning that underlies the acquisition of motor performance. Inspired by neurophysiological findings showing that cerebellar LTD and cortical LTP would both be disrupted by sub-100ms latencies, we hypothesized that implicit sensorimotor learning may be particularly sensitive to these short latencies. Remarkably, we find that improving latency by just 60ms, from 85 to 25ms in latency-optimized experiments, increases implicit learning by 50% and proportionally decreases explicit learning, resulting in a dramatic reorganization of sensorimotor memory. We go on to show that implicit sensorimotor learning is considerably more sensitive to latencies in the sub-100ms range than at higher latencies, in line with the latency-specific neural plasticity that has been observed. This suggests a clear benefit for latency reduction in computer-based training that involves implicit sensorimotor learning and that across-study differences in implicit motor learning might often be explained by disparities in feedback latency.

Large-scale citizen science reveals predictors of sensorimotor adaptation

Sensorimotor adaptation is essential for keeping our movements well calibrated in response to changes in the body and environment. For over a century, researchers have studied sensorimotor adaptation in laboratory settings that typically involve small sample sizes. While this approach has proved useful for characterizing different learning processes, laboratory studies are not well suited for exploring the myriad of factors that may modulate human performance. Here, using a citizen science website, we collected over 2,000 sessions of data on a visuomotor rotation task. This unique dataset has allowed us to replicate, reconcile and challenge classic findings in the learning and memory literature, as well as discover unappreciated demographic constraints associated with implicit and explicit processes that support sensorimotor adaptation. More generally, this study exemplifies how a large-scale exploratory approach can complement traditional hypothesis-driven laboratory research in advancing sensorimotor neuroscience.

NSF DARE-transforming modeling in neurorehabilitation: a patient-in-the-loop framework

In 2023, the National Science Foundation (NSF) and the National Institute of Health (NIH) brought together engineers, scientists, and clinicians by sponsoring a conference on computational modelling in neurorehabiilitation. To facilitate multidisciplinary collaborations and improve patient care, in this perspective piece we identify where and how computational modelling can support neurorehabilitation. To address the where, we developed a patient-in-the-loop framework that uses multiple and/or continual measurements to update diagnostic and treatment model parameters, treatment type, and treatment prescription, with the goal of maximizing clinically-relevant functional outcomes. This patient-in-the-loop framework has several key features: (i) it includes diagnostic and treatment models, (ii) it is clinically-grounded with the International Classification of Functioning, Disability and Health (ICF) and patient involvement, (iii) it uses multiple or continual data measurements over time, and (iv) it is applicable to a range of neurological and neurodevelopmental conditions. To address the how, we identify state-of-the-art and highlight promising avenues of future research across the realms of sensorimotor adaptation, neuroplasticity, musculoskeletal, and sensory & pain computational modelling. We also discuss both the importance of and how to perform model validation, as well as challenges to overcome when implementing computational models within a clinical setting. The patient-in-the-loop approach offers a unifying framework to guide multidisciplinary collaboration between computational and clinical stakeholders in the field of neurorehabilitation.

A software tool for at-home measurement of sensorimotor adaptation

Sensorimotor adaptation is traditionally studied in well-controlled laboratory settings with specialized equipment. However, recent public health concerns such as the COVID-19 pandemic, as well as a desire to recruit a more diverse study population, have led the motor control community to consider at-home study designs. At-home motor control experiments are still rare because of the requirement to write software that can be easily used by anyone on any platform. To this end, we developed software that runs locally on a personal computer. The software provides audiovisual instructions and measures the ability of the subject to control the cursor in the context of visuomotor perturbations. We tested the software on a group of at-home participants and asked whether the adaptation principles inferred from in-lab measurements were reproducible in the at-home setting. For example, we manipulated the perturbations to test whether there were changes in adaptation rates (savings and interference), whether adaptation was associated with multiple timescales of memory (spontaneous recovery), and whether we could selectively suppress subconscious learning (delayed feedback, perturbation variability) or explicit strategies (limited reaction time). We found remarkable similarity between in-lab and at-home behaviors across these experimental conditions. Thus, we developed a software tool that can be used by research teams with little or no programming experience to study mechanisms of adaptation in an at-home setting.

Explicit learning based on reward prediction error facilitates agile motor adaptations

Error based motor learning can be driven by both sensory prediction error and reward prediction error. Learning based on sensory prediction error is termed sensorimotor adaptation, while learning based on reward prediction error is termed reward learning. To investigate the characteristics and differences between sensorimotor adaptation and reward learning, we adapted a visuomotor paradigm where subjects performed arm movements while presented with either the sensory prediction error, signed end-point error, or binary reward. Before each trial, perturbation indicators in the form of visual cues were presented to inform the subjects of the presence and direction of the perturbation. To analyse the interconnection between sensorimotor adaptation and reward learning, we designed a computational model that distinguishes between the two prediction errors. Our results indicate that subjects adapted to novel perturbations irrespective of the type of prediction error they received during learning, and they converged towards the same movement patterns. Sensorimotor adaptations led to a pronounced aftereffect, while adaptation based on reward consequences produced smaller aftereffects suggesting that reward learning does not alter the internal model to the same degree as sensorimotor adaptation. Even though all subjects had learned to counteract two different perturbations separately, only those who relied on explicit learning using reward prediction error could timely adapt to the randomly changing perturbation. The results from the computational model suggest that sensorimotor and reward learning operate through distinct adaptation processes and that only sensorimotor adaptation changes the internal model, whereas reward learning employs explicit strategies that do not result in aftereffects. Additionally, we demonstrate that when humans learn motor tasks, they utilize both learning processes to successfully adapt to the new environments.

Neurophysiological evidence of sensory prediction errors driving speech sensorimotor adaptation

The human sensorimotor system has a remarkable ability to quickly and efficiently learn movements from sensory experience. A prominent example is sensorimotor adaptation, learning that characterizes the sensorimotor system's response to persistent sensory errors by adjusting future movements to compensate for those errors. Despite being essential for maintaining and fine-tuning motor control, mechanisms underlying sensorimotor adaptation remain unclear. A component of sensorimotor adaptation is implicit (i.e., the learner is unaware of the learning process) which has been suggested to result from sensory prediction errors-the discrepancies between predicted sensory consequences of motor commands and actual sensory feedback. However, to date no direct neurophysiological evidence that sensory prediction errors drive adaptation has been demonstrated. Here, we examined prediction errors via magnetoencephalography (MEG) imaging of the auditory cortex during sensorimotor adaptation of speech to altered auditory feedback, an entirely implicit adaptation task. Specifically, we measured how speaking-induced suppression (SIS)--a neural representation of auditory prediction errors--changed over the trials of the adaptation experiment. SIS refers to the suppression of auditory cortical response to speech onset (in particular, the M100 response) to self-produced speech when compared to the response to passive listening to identical playback of that speech. SIS was reduced (reflecting larger prediction errors) during the early learning phase compared to the initial unaltered feedback phase. Furthermore, reduction in SIS positively correlated with behavioral adaptation extents, suggesting that larger prediction errors were associated with more learning. In contrast, such a reduction in SIS was not found in a control experiment in which participants heard unaltered feedback and thus did not adapt. In addition, in some participants who reached a plateau in the late learning phase, SIS increased (reflecting smaller prediction errors), demonstrating that prediction errors were minimal when there was no further adaptation. Together, these findings provide the first neurophysiological evidence for the hypothesis that prediction errors drive human sensorimotor adaptation.

Strategy-based motor learning decreases the post-movement β power

Motor learning depends on the joint contribution of several processes including cognitive strategies aiming at goal achievement and prediction error-driven implicit adaptation. Understanding this functional interplay and its clinical implications requires insight into the individual learning processes, including at a neural level. Here, we set out to examine the impact of learning a cognitive strategy, over and above implicit adaptation, on the oscillatory post-movement β rebound (PMBR), which typically decreases in power following (visuo)motor perturbations. Healthy participants performed reaching movements towards a target, with online visual feedback replacing the view of their moving hand. The feedback was sometimes rotated, either relative to their movements (visuomotor rotation) or invariant to their movements (and relative to the target; clamped feedback), always for two consecutive trials interspersed between non-rotated trials. In both conditions, the first trial with a rotation was unpredictable. On the second trial, the task was either to re-aim, and thereby compensate for the rotation experienced in the first trial (visuomotor rotation; Compensate condition), or to ignore the rotation and keep on aiming at the target (clamped feedback; Ignore condition). After-effects did not differ between conditions, indicating that the amount of implicit learning was similar, while large differences in movement direction in the second rotated trial between conditions indicated that participants successfully acquired re-aiming strategies. Importantly, PMBR power following the first rotated trial was modulated differently in the two conditions. Specifically, it decreased in both conditions, but this effect was larger when participants had to acquire a cognitive strategy and prepare to re-aim. Our results therefore suggest that the PMBR is modulated by cognitive demands of motor learning, possibly reflecting the evaluation of a behaviourally significant goal achievement error.

Sensory prediction error drives subconscious motor learning outside of the laboratory

Sensorimotor adaptation is supported by at least two parallel learning systems: an intentionally controlled explicit strategy and an involuntary implicit learning system. Past work focused on constrained reaches or finger movements in laboratory environments has shown subconscious learning systems to be driven in part by sensory prediction error (SPE), i.e., the mismatch between the realized and expected outcome of an action. We designed a ball rolling task to explore whether SPEs can drive implicit motor adaptation during complex whole body movements that impart physical motion on external objects. After applying a visual shift, participants rapidly adapted their rolling angles to reduce the error between the ball and the target. We removed all visual feedback and told participants to aim their throw directly toward the primary target, revealing an unintentional 5.06° implicit adjustment to reach angles that decayed over time. To determine whether this implicit adaptation was driven by SPE, we gave participants a second aiming target that would "solve" the visual shift, as in the study by Mazzoni and Krakauer (Mazzoni P, Krakauer JW. J Neurosci 26: 3642-3645, 2006). Remarkably, after rapidly reducing ball-rolling error to zero (due to enhancements in strategic aiming), the additional aiming target caused rolling angles to deviate beyond the primary target by 3.15°. This involuntary overcompensation, which worsened task performance, is a hallmark of SPE-driven implicit learning. These results show that SPE-driven implicit processes, previously observed within simplified finger or planar reaching movements, actively contribute to motor adaptation in more complex naturalistic skill-based tasks.NEW & NOTEWORTHY Implicit and explicit learning systems have been detected using simple, constrained movements inside the laboratory. How these systems impact movements during complex whole body, skill-based tasks has not been established. Here, we demonstrate that sensory prediction errors significantly impact how a person updates their movements, replicating findings from the laboratory in an unconstrained ball-rolling task. This real-world validation is an important step toward explaining how subconscious learning helps humans execute common motor skills in dynamic environments.

Mechanisms of sensorimotor adaptation in a hierarchical state feedback control model of speech

Upon perceiving sensory errors during movements, the human sensorimotor system updates future movements to compensate for the errors, a phenomenon called sensorimotor adaptation. One component of this adaptation is thought to be driven by sensory prediction errors-discrepancies between predicted and actual sensory feedback. However, the mechanisms by which prediction errors drive adaptation remain unclear. Here, auditory prediction error-based mechanisms involved in speech auditory-motor adaptation were examined via the feedback aware control of tasks in speech (FACTS) model. Consistent with theoretical perspectives in both non-speech and speech motor control, the hierarchical architecture of FACTS relies on both the higher-level task (vocal tract constrictions) as well as lower-level articulatory state representations. Importantly, FACTS also computes sensory prediction errors as a part of its state feedback control mechanism, a well-established framework in the field of motor control. We explored potential adaptation mechanisms and found that adaptive behavior was present only when prediction errors updated the articulatory-to-task state transformation. In contrast, designs in which prediction errors updated forward sensory prediction models alone did not generate adaptation. Thus, FACTS demonstrated that 1) prediction errors can drive adaptation through task-level updates, and 2) adaptation is likely driven by updates to task-level control rather than (only) to forward predictive models. Additionally, simulating adaptation with FACTS generated a number of important hypotheses regarding previously reported phenomena such as identifying the source(s) of incomplete adaptation and driving factor(s) for changes in the second formant frequency during adaptation to the first formant perturbation. The proposed model design paves the way for a hierarchical state feedback control framework to be examined in the context of sensorimotor adaptation in both speech and non-speech effector systems.

Whole-brain dynamics of human sensorimotor adaptation

Humans vary greatly in their motor learning abilities, yet little is known about the neural processes that underlie this variability. We identified distinct profiles of human sensorimotor adaptation that emerged across 2 days of learning, linking these profiles to the dynamics of whole-brain functional networks early on the first day when cognitive strategies toward sensorimotor adaptation are believed to be most prominent. During early learning, greater recruitment of a network of higher-order brain regions, involving prefrontal and anterior temporal cortex, was associated with faster learning. At the same time, greater integration of this "cognitive network" with a sensorimotor network was associated with slower learning, consistent with the notion that cognitive strategies toward adaptation operate in parallel with implicit learning processes of the sensorimotor system. On the second day, greater recruitment of a network that included the hippocampus was associated with faster learning, consistent with the notion that declarative memory systems are involved with fast relearning of sensorimotor mappings. Together, these findings provide novel evidence for the role of higher-order brain systems in driving variability in adaptation.

Adapting to visuomotor rotations in stepped increments increases implicit motor learning

Human motor adaptation relies on both explicit conscious strategies and implicit unconscious updating of internal models to correct motor errors. Implicit adaptation is powerful, requiring less preparation time before executing adapted movements, but recent work suggests it is limited to some absolute magnitude regardless of the size of a visuomotor perturbation when the perturbation is introduced abruptly. It is commonly assumed that gradually introducing a perturbation should lead to improved implicit learning beyond this limit, but outcomes are conflicting. We tested whether introducing a perturbation in two distinct gradual methods can overcome the apparent limit and explain past conflicting findings. We found that gradually introducing a perturbation in a stepped manner, where participants were given time to adapt to each partial step before being introduced to a larger partial step, led to ~ 80% higher implicit aftereffects of learning, but introducing it in a ramped manner, where participants adapted larger rotations on each subsequent reach, did not. Our results clearly show that gradual introduction of a perturbation can lead to substantially larger implicit adaptation, as well as identify the type of introduction that is necessary to do so.

Implicit sensorimotor adaptation is preserved in Parkinson's disease

Our ability to enact successful goal-directed actions involves multiple learning processes. Among these processes, implicit motor adaptation ensures that the sensorimotor system remains finely tuned in response to changes in the body and environment. Whether Parkinson's disease impacts implicit motor adaptation remains a contentious area of research: whereas multiple reports show impaired performance in this population, many others show intact performance. While there is a range of methodological differences across studies, one critical issue is that performance in many of the studies may reflect a combination of implicit adaptation and strategic re-aiming. Here, we revisited this controversy using a visuomotor task designed to isolate implicit adaptation. In two experiments, we found that adaptation in response to a wide range of visual perturbations was similar in Parkinson's disease and matched control participants. Moreover, in a meta-analysis of previously published and unpublished work, we found that the mean effect size contrasting Parkinson's disease and controls across 16 experiments involving over 200 participants was not significant. Together, these analyses indicate that implicit adaptation is preserved in Parkinson's disease, offering a fresh perspective on the role of the basal ganglia in sensorimotor learning.

Neural excursions from manifold structure explain patterns of learning during human sensorimotor adaptation

Humans vary greatly in their motor learning abilities, yet little is known about the neural mechanisms that underlie this variability. Recent neuroimaging and electrophysiological studies demonstrate that large-scale neural dynamics inhabit a low-dimensional subspace or manifold, and that learning is constrained by this intrinsic manifold architecture. Here, we asked, using functional MRI, whether subject-level differences in neural excursion from manifold structure can explain differences in learning across participants. We had subjects perform a sensorimotor adaptation task in the MRI scanner on 2 consecutive days, allowing us to assess their learning performance across days, as well as continuously measure brain activity. We find that the overall neural excursion from manifold activity in both cognitive and sensorimotor brain networks is associated with differences in subjects' patterns of learning and relearning across days. These findings suggest that off-manifold activity provides an index of the relative engagement of different neural systems during learning, and that subject differences in patterns of learning and relearning are related to reconfiguration processes occurring in cognitive and sensorimotor networks.

Interactions between sensory prediction error and task error during implicit motor learning

Implicit motor recalibration allows us to flexibly move in novel and changing environments. Conventionally, implicit recalibration is thought to be driven by errors in predicting the sensory outcome of movement (i.e., sensory prediction errors). However, recent studies have shown that implicit recalibration is also influenced by errors in achieving the movement goal (i.e., task errors). Exactly how sensory prediction errors and task errors interact to drive implicit recalibration and, in particular, whether task errors alone might be sufficient to drive implicit recalibration remain unknown. To test this, we induced task errors in the absence of sensory prediction errors by displacing the target mid-movement. We found that task errors alone failed to induce implicit recalibration. In additional experiments, we simultaneously varied the size of sensory prediction errors and task errors. We found that implicit recalibration driven by sensory prediction errors could be continuously modulated by task errors, revealing an unappreciated dependency between these two sources of error. Moreover, implicit recalibration was attenuated when the target was simply flickered in its original location, even though this manipulation did not affect task error - an effect likely attributed to attention being directed away from the feedback cursor. Taken as a whole, the results were accounted for by a computational model in which sensory prediction errors and task errors, modulated by attention, interact to determine the extent of implicit recalibration.

Task space exploration improves adaptation after incompatible virtual surgeries

Humans have a remarkable capacity to learn new motor skills, a process that requires novel muscle activity patterns. Muscle synergies may simplify the generation of muscle patterns through the selection of a small number of synergy combinations. Learning new motor skills may then be achieved by acquiring novel muscle synergies. In a previous study, we used myoelectric control to construct virtual surgeries that altered the mapping from muscle activity to cursor movements. After compatible virtual surgeries, which could be compensated by recombining subject-specific muscle synergies, participants adapted quickly. In contrast, after incompatible virtual surgeries, which could not be compensated by recombining existing synergies, participants explored new muscle patterns, but failed to adapt. Here, we tested whether task space exploration can promote learning of novel muscle synergies, required to overcome an incompatible surgery. Participants performed the same reaching task as in our previous study, but with more time to complete each trial, thus allowing for exploration. We found an improvement in trial success after incompatible virtual surgeries. Remarkably, improvements in movement direction accuracy after incompatible surgeries occurred faster for corrective movements than for the initial movement, suggesting that learning of new synergies is more effective when used for feedback control. Moreover, reaction time was significantly higher after incompatible than after compatible virtual surgeries, suggesting an increased use of an explicit adaptive strategy to overcome incompatible surgeries. Taken together, these results indicate that exploration is important for skill learning and suggest that human participants, with sufficient time, can learn new muscle synergies.

Competition between parallel sensorimotor learning systems

Sensorimotor learning is supported by at least two parallel systems: a strategic process that benefits from explicit knowledge and an implicit process that adapts subconsciously. How do these systems interact? Does one system's contributions suppress the other, or do they operate independently? Here, we illustrate that during reaching, implicit and explicit systems both learn from visual target errors. This shared error leads to competition such that an increase in the explicit system's response siphons away resources that are needed for implicit adaptation, thus reducing its learning. As a result, steady-state implicit learning can vary across experimental conditions, due to changes in strategy. Furthermore, strategies can mask changes in implicit learning properties, such as its error sensitivity. These ideas, however, become more complex in conditions where subjects adapt using multiple visual landmarks, a situation which introduces learning from sensory prediction errors in addition to target errors. These two types of implicit errors can oppose each other, leading to another type of competition. Thus, during sensorimotor adaptation, implicit and explicit learning systems compete for a common resource: error.

Mechanisms of Human Motor Learning Do Not Function Independently

Human motor learning is governed by a suite of interacting mechanisms each one of which modifies behavior in distinct ways and rely on different neural circuits. In recent years, much attention has been given to one type of motor learning, called motor adaptation. Here, the field has generally focused on the interactions of three mechanisms: sensory prediction error SPE-driven, explicit (strategy-based), and reinforcement learning. Studies of these mechanisms have largely treated them as modular, aiming to model how the outputs of each are combined in the production of overt behavior. However, when examined closely the results of some studies also suggest the existence of additional interactions between the sub-components of each learning mechanism. In this perspective, we propose that these sub-component interactions represent a critical means through which different motor learning mechanisms are combined to produce movement; understanding such interactions is critical to advancing our knowledge of how humans learn new behaviors. We review current literature studying interactions between SPE-driven, explicit, and reinforcement mechanisms of motor learning. We then present evidence of sub-component interactions between SPE-driven and reinforcement learning as well as between SPE-driven and explicit learning from studies of people with cerebellar degeneration. Finally, we discuss the implications of interactions between learning mechanism sub-components for future research in human motor learning.

Evidence that endpoint feedback facilitates intermanual transfer of visuomotor force learning by a cognitive strategy

Humans continuously adapt their movement to a novel environment by recalibrating their sensorimotor system. Recent evidence, however, shows that explicit planning to compensate for external changes, i.e., a cognitive strategy, can also aid performance. If such a strategy is planned in external space, it should improve performance in an effector-independent manner. We tested this hypothesis by examining whether promoting a cognitive strategy during a visual-force adaptation task performed in one hand can facilitate learning for the opposite hand. Participants rapidly adjusted the height of visual bar on screen to a target level by isometrically exerting force on a handle using their right hand. Visuomotor gain increased during the task and participants learned the increased gain. Visual feedback was continuously provided for one group, whereas for another group only the endpoint of the force trajectory was presented. The latter has been reported to promote cognitive strategy use. We found that endpoint feedback produced stronger intermanual transfer of learning and slower response times than continuous feedback. In a separate experiment, we found evidence that aftereffects are reduced when only endpoint feedback is provided, a finding that has been consistently observed when cognitive strategies are used. The results suggest that intermanual transfer can be facilitated by a cognitive strategy. This indicates that the behavioral observation of intermanual transfer can be achieved either by forming an effector-independent motor representation or by sharing an effector-independent cognitive strategy between the hands.

NEW & NOTEWORTHY The causes and consequences of cognitive strategy use are poorly understood. We tested whether a visuomotor task learned in a manner that may promote cognitive strategy use causes greater generalization across effectors. Visual feedback was manipulated to promote cognitive strategy use. Learning consistent with cognitive strategy use for one hand transferred to the unlearned hand. Our result suggests that intermanual transfer can result from a common cognitive strategy used to control both hands.

Human Variation in Error-Based and Reinforcement Motor Learning Is Associated With Entorhinal Volume

Error-based and reward-based processes are critical for motor learning and are thought to be mediated via distinct neural pathways. However, recent behavioral work in humans suggests that both learning processes can be bolstered by the use of cognitive strategies, which may mediate individual differences in motor learning ability. It has been speculated that medial temporal lobe regions, which have been shown to support motor sequence learning, also support the use of cognitive strategies in error-based and reinforcement motor learning. However, direct evidence in support of this idea remains sparse. Here we first show that better overall learning during error-based visuomotor adaptation is associated with better overall learning during the reward-based shaping of reaching movements. Given the cognitive contribution to learning in both of these tasks, these results support the notion that strategic processes, associated with better performance, drive intersubject variation in both error-based and reinforcement motor learning. Furthermore, we show that entorhinal cortex volume is larger in better learning individuals-characterized across both motor learning tasks-compared with their poorer learning counterparts. These results suggest that individual differences in learning performance during error and reinforcement learning are related to neuroanatomical differences in entorhinal cortex.

Revisiting the Role of the Medial Temporal Lobe in Motor Learning

Classic taxonomies of memory distinguish explicit and implicit memory systems, placing motor skills squarely in the latter branch. This assertion is in part a consequence of foundational discoveries showing significant motor learning in amnesics. Those findings suggest that declarative memory processes in the medial temporal lobe (MTL) do not contribute to motor learning. Here, we revisit this issue, testing an individual (L. S. J.) with severe MTL damage on four motor learning tasks and comparing her performance to age-matched controls. Consistent with previous findings in amnesics, we observed that L. S. J. could improve motor performance despite having significantly impaired declarative memory. However, she tended to perform poorly relative to age-matched controls, with deficits apparently related to flexible action selection. Further supporting an action selection deficit, L. S. J. fully failed to learn a task that required the acquisition of arbitrary action-outcome associations. We thus propose a modest revision to the classic taxonomic model: Although MTL-dependent memory processes are not necessary for some motor learning to occur, they play a significant role in the acquisition, implementation, and retrieval of action selection strategies. These findings have implications for our understanding of the neural correlates of motor learning, the psychological mechanisms of skill, and the theory of multiple memory systems.

Separation of multiple motor memories through implicit and explicit processes

Acquisition of multiple motor skills without interference is a remarkable ability in daily life. During adaptation to opposing perturbations, a common paradigm to study this ability, each perturbation can be successfully learned when a contextual follow-through movement is associated with the direction of the perturbation. It is still unclear, however, to what extent this learning engages the cognitive explicit process and the implicit process. Here, we untangled the individual contributions of the explicit and implicit components while participants learned opposing visuomotor perturbations, with a second unperturbed follow-through movement. In Exp. 1 we replicated previous adaptation results and showed that follow-through movements also allow learning for opposing visuomotor rotations. For one group of participants in Exp. 2 we isolated strategic explicit learning, while for another group we isolated the implicit component. Our data showed that opposing perturbations could be fully learned by explicit strategies; but when strategy was restricted, distinct implicit processes contributed to learning. In Exp.3, we examined whether learning is influenced by the disparity between the follow-through contexts. We found that the location of follow-through targets had little effect on total learning, yet it led to more instances in which participants failed to learn the task. In Exp. 4, we explored the generalization capability to untrained targets. Participants showed near-flat generalization of the implicit and explicit processes. Overall, our results indicate that follow-through contextual cues might activate, in part, top-down cognitive factors that influence not only the dynamics of the explicit learning, but also the implicit process.

Contextual inference underlies the learning of sensorimotor repertoires

ASBTRACT

Humans spend a lifetime learning, storing and refining a repertoire of motor memories. For example, through experience, we become proficient at manipulating a large range of objects with distinct dynamical properties. However, it is unknown what principle underlies how our continuous stream of sensorimotor experience is segmented into separate memories and how we adapt and use this growing repertoire. Here we develop a theory of motor learning based on the key principle that memory creation, updating and expression are all controlled by a single computation-contextual inference. Our theory reveals that adaptation can arise both by creating and updating memories (proper learning) and by changing how existing memories are differentially expressed (apparent learning). This insight enables us to account for key features of motor learning that had no unified explanation: spontaneous recovery¹, savings², anterograde interference³, how environmental consistency affects learning rate^4,5 and the distinction between explicit and implicit learning⁶. Critically, our theory also predicts new phenomena-evoked recovery and context-dependent single-trial learning-which we confirm experimentally. These results suggest that contextual inference, rather than classical single-context mechanisms^1,4,7-9, is the key principle underlying how a diverse set of experiences is reflected in our motor behaviour.

The neural mechanisms of manual dexterity

The hand endows us with unparalleled precision and versatility in our interactions with objects, from mundane activities such as grasping to extraordinary ones such as virtuoso pianism. The complex anatomy of the human hand combined with expansive and specialized neuronal control circuits allows a wide range of precise manual behaviours. To support these behaviours, an exquisite sensory apparatus, spanning the modalities of touch and proprioception, conveys detailed and timely information about our interactions with objects and about the objects themselves. The study of manual dexterity provides a unique lens into the sensorimotor mechanisms that endow the nervous system with the ability to flexibly generate complex behaviour.

Brain oscillatory correlates of visuomotor adaptive learning

Sensorimotor adaptation involves the recalibration of the mapping between motor command and sensory feedback in response to movement errors. Although adaptation operates within individual movements on a trial-to-trial basis, it can also undergo learning when adaptive responses improve over the course of many trials. Brain oscillatory activities related to these "adaptation" and "learning" processes remain unclear. The main reason for this is that previous studies principally focused on the beta band, which confined the outcome message to trial-to-trial adaptation. To provide a wider understanding of adaptive learning, we decoded visuomotor tasks with constant, random or no perturbation from EEG recordings in different bandwidths and brain regions using a multiple kernel learning approach. These different experimental tasks were intended to separate trial-to-trial adaptation from the formation of the new visuomotor mapping across trials. We found changes in EEG power in the post-movement period during the course of the visuomotor-constant rotation task, in particular an increased (i) theta power in prefrontal region, (ii) beta power in supplementary motor area, and (iii) gamma power in motor regions. Classifying the visuomotor task with constant rotation versus those with random or no rotation, we were able to relate power changes in beta band mainly to trial-to-trial adaptation to error while changes in theta band would relate rather to the learning of the new mapping. Altogether, this suggested that there is a tight relationship between modulation of the synchronization of low (theta) and higher (essentially beta) frequency oscillations in prefrontal and sensorimotor regions, respectively, and adaptive learning.

Normal Aging Affects the Short-Term Temporal Stability of Implicit, But Not Explicit, Motor Learning following Visuomotor Adaptation

Normal aging is associated with a decline in memory and motor learning ability. However, the exact form of these impairments (e.g., the short-term temporal stability and affected learning mechanisms) is largely unknown. Here, we used a sensorimotor adaptation task to examine changes in the temporal stability of two forms of learning (explicit and implicit) because of normal aging. Healthy young subjects (age range, 19–28 years; 20 individuals) and older human subjects (age range, 63–85 years; 19 individuals) made reaching movements in response to altered visual feedback. On each trial, subjects turned a rotation dial to select an explicit aiming direction. Once selected, the display was removed and subjects moved the cursor from the start position to the target. After initial training with the rotational feedback perturbation, subjects completed a series of probe trials at different delay periods to systematically assess the short-term retention of learning. For both groups, the explicit aiming showed no significant decrease over 1.5 min. However, this was not the case for implicit learning; the decay pattern was markedly different between groups. Older subjects showed a linear decrease of the implicit component of adaptation over time, while young subjects showed an exponential decay over the same period (time constant, 25.61 s). Although older subjects adapted at a similar rate, these results suggest natural aging selectively impacts the short-term (seconds to minutes) temporal stability of implicit motor learning mechanisms. This understanding may provide a means to dissociate natural aging memory impairments from deficits caused by brain disorders that progress with aging.

Adaptive control of movement deceleration during saccades

As you read this text, your eyes make saccades that guide your fovea from one word to the next. Accuracy of these movements require the brain to monitor and learn from visual errors. A current model suggests that learning is supported by two different adaptive processes, one fast (high error sensitivity, low retention), and the other slow (low error sensitivity, high retention). Here, we searched for signatures of these hypothesized processes and found that following experience of a visual error, there was an adaptive change in the motor commands of the subsequent saccade. Surprisingly, this adaptation was not uniformly expressed throughout the movement. Rather, after experience of a single error, the adaptive response in the subsequent trial was limited to the deceleration period. After repeated exposure to the same error, the acceleration period commands also adapted, and exhibited resistance to forgetting during set-breaks. In contrast, the deceleration period commands adapted more rapidly, but suffered from poor retention during these same breaks. State-space models suggested that acceleration and deceleration periods were supported by a shared adaptive state which re-aimed the saccade, as well as two separate processes which resembled a two-state model: one that learned slowly and contributed primarily via acceleration period commands, and another that learned rapidly but contributed primarily via deceleration period commands.

An implicit memory of errors limits human sensorimotor adaptation

During extended motor adaptation, learning appears to saturate despite persistence of residual errors. This adaptation limit is not fixed but varies with perturbation variance; when variance is high, residual errors become larger. These changes in total adaptation could relate to either implicit or explicit learning systems. Here, we found that when adaptation relied solely on the explicit system, residual errors disappeared and learning was unaltered by perturbation variability. In contrast, when learning depended entirely, or in part, on implicit learning, residual errors reappeared. Total implicit adaptation decreased in the high-variance environment due to changes in error sensitivity, not in forgetting. These observations suggest a model in which the implicit system becomes more sensitive to errors when they occur in a consistent direction. Thus, residual errors in motor adaptation are at least in part caused by an implicit learning system that modulates its error sensitivity in response to the consistency of past errors.

Measures of explicit and implicit in motor learning: what we know and what we don't

Adaptation tasks are a key tool in characterizing the contribution of explicit and implicit processes to sensorimotor learning. However, different assumptions and ideas underlie methods used to measure these processes, leading to inconsistencies between studies. For instance, it is still unclear explicit and implicit combine additively. Cognitive studies of explicit and implicit processes show how non-additivity and bias in measurement can distort results. We argue that to understand explicit and implicit processes in visuomotor adaptation, we need a stronger characterization of the phenomenology and a richer set of models to test it on.

De novo learning versus adaptation of continuous control in a manual tracking task

How do people learn to perform tasks that require continuous adjustments of motor output, like riding a bicycle? People rely heavily on cognitive strategies when learning discrete movement tasks, but such time-consuming strategies are infeasible in continuous control tasks that demand rapid responses to ongoing sensory feedback. To understand how people can learn to perform such tasks without the benefit of cognitive strategies, we imposed a rotation/mirror reversal of visual feedback while participants performed a continuous tracking task. We analyzed behavior using a system identification approach, which revealed two qualitatively different components of learning: adaptation of a baseline controller and formation of a new, task-specific continuous controller. These components exhibited different signatures in the frequency domain and were differentially engaged under the rotation/mirror reversal. Our results demonstrate that people can rapidly build a new continuous controller de novo and can simultaneously deploy this process with adaptation of an existing controller.

Increasing the gradient of energetic cost does not initiate adaptation in human walking

When in a new situation, the nervous system may benefit from adapting its control policy. In determining whether or not to initiate this adaptation, the nervous system may rely on some features of the new situation. Here, we tested whether one such feature is salient cost savings. We changed cost saliency by manipulating the gradient of participants' energetic cost landscape during walking. We hypothesized that steeper gradients would cause participants to spontaneously adapt their step frequency to lower costs. To manipulate the gradient, a mechatronic system applied controlled fore-aft forces to the waist of participants as a function of their step frequency as they walked on a treadmill. These forces increased the energetic cost of walking at high step frequencies and reduced it at low step frequencies. We successfully created three cost landscapes of increasing gradients, where the natural variability in participants' step frequency provided cost changes of 3.6% (shallow), 7.2% (intermediate), and 10.2% (steep). Participants did not spontaneously initiate adaptation in response to any of the gradients. Using metronome-guided walking-a previously established protocol for eliciting initiation of adaptation-participants next experienced a step frequency with a lower cost. Participants then adapted by -1.41 ± 0.81 (P = 0.007) normalized units away from their originally preferred step frequency obtaining cost savings of 4.80% ± 3.12%. That participants would adapt under some conditions, but not in response to steeper cost gradients, suggests that the nervous system does not solely rely on the gradient of energetic cost to initiate adaptation in novel situations.NEW & NOTEWORTHY People can adapt to novel conditions but often require cues to initiate the adaptation. Using a mechatronic system to reshape energetic cost gradients during treadmill walking, we tested whether the nervous system can use information present in the cost gradient to spontaneously initiate adaptation. We found that our participants did not spontaneously initiate adaptation even in the steepest gradient. The nervous system does not rely solely on the cost gradient when initiating adaptation.

Taking aim at the perceptual side of motor learning: exploring how explicit and implicit learning encode perceptual error information through depth vision

Motor learning in visuomotor adaptation tasks results from both explicit and implicit processes, each responding differently to an error signal. Although the motor output side of these processes has been extensively studied, the visual input side is relatively unknown. We investigated if and how depth perception affects the computation of error information by explicit and implicit motor learning. Two groups of participants made reaching movements to bring a virtual cursor to a target in the frontoparallel plane. The Delayed group was allowed to reaim and their feedback was delayed to emphasize explicit learning, whereas the camped group received task-irrelevant clamped cursor feedback and continued to aim straight at the target to emphasize implicit adaptation. Both groups played this game in a highly detailed virtual environment (depth condition), leveraging a cover task of playing darts in a virtual tavern, and in an empty environment (no-depth condition). The delayed group showed an increase in error sensitivity under depth relative to no-depth. In contrast, the clamped group adapted to the same degree under both conditions. The movement kinematics of the delayed participants also changed under the depth condition, consistent with the target appearing more distant, unlike the Clamped group. A comparison of the delayed behavioral data with a perceptual task from the same individuals showed that the greater reaiming in the depth condition was consistent with an increase in the scaling of the error distance and size. These findings suggest that explicit and implicit learning processes may rely on different sources of perceptual information.NEW & NOTEWORTHY We leveraged a classic sensorimotor adaptation task to perform a first systematic assessment of the role of perceptual cues in the estimation of an error signal in the 3-D space during motor learning. We crossed two conditions presenting different amounts of depth information, with two manipulations emphasizing explicit and implicit learning processes. Explicit learning responded to the visual conditions, consistent with perceptual reports, whereas implicit learning appeared to be independent of them.

Hierarchical motor adaptations negotiate failures during force field learning

Humans have the amazing ability to learn the dynamics of the body and environment to develop motor skills. Traditional motor studies using arm reaching paradigms have viewed this ability as the process of ‘internal model adaptation’. However, the behaviors have not been fully explored in the case when reaches fail to attain the intended target. Here we examined human reaching under two force fields types; one that induces failures (i.e., target errors), and the other that does not. Our results show the presence of a distinct failure-driven adaptation process that enables quick task success after failures, and before completion of internal model adaptation, but that can result in persistent changes to the undisturbed trajectory. These behaviors can be explained by considering a hierarchical interaction between internal model adaptation and the failure-driven adaptation of reach direction. Our findings suggest that movement failure is negotiated using hierarchical motor adaptations by humans.

How do we improve actions after a movement failure? Although negotiating movement failures is obviously crucial, previous motor-control studies have predominantly examined human movement adaptations in the absence of failures, and it remains unclear how failures affect subsequent movement adaptations. Here we examined this issue by developing a novel force field adaptation task where the hand movement during an arm reaching is perturbed by novel forces that induce a large target error, that is a failure. Our experimental observation and computational modeling show that, in addition to the popular ‘internal model learning’ process of motor adaptations, humans also utilize a ‘failure-negotiating’ process, that enables them to quickly improve movements in the presence of failure, even at the expense of increased arm trajectory deflections, which are subsequently reduced gradually with training after the achievement of the task success. Our results suggest that a hierarchical interaction between these two processes is a key for humans to negotiate movement failures.

Loss of peripheral vestibular input alters the statistics of head movement experienced during natural self-motion

KEY POINTS

Sensory systems are adapted to the statistical structure of natural stimuli, thereby optimizing neural coding. Head motion during natural activities is first sensed and then processed by central vestibulo-motor pathways to influence subsequent behaviour, thereby establishing a feedback loop. To investigate the role of this vestibular feedback on the statistical structure of the head movements, we compared head movements in patients with unilateral vestibular loss and healthy controls. We show that the loss of vestibular feedback substantially alters the statistical structure of head motion for activities that require rapid online feedback control and predict this change by modelling the effects of increased movement variability. Our findings suggest that, following peripheral vestibular loss, changes in the reliability of the sensory input to central pathways impact the statistical structure of head motion during voluntary behaviours.

ABSTRACT

It is widely believed that sensory systems are adapted to optimize neural coding of their natural stimuli. Recent evidence suggests that this is the case for the vestibular system, which senses head movement and contributes to essential functions ranging from the most automatic reflexes to voluntary motor control. During everyday behaviours, head motion is sensed by the vestibular system. In turn, this sensory feedback influences subsequent behaviour, raising the questions of whether and how real-time feedback provided by the vestibular system alters the statistical structure of head movements. We predicted that a reduction in vestibular feedback would alter head movement statistics, particularly for tasks reliant on rapid vestibular feedback. To test this proposal, we recorded six-dimensional head motion in patients with variable degrees of unilateral vestibular loss during standard balance and gait tasks, as well as dynamic self-paced activities. While distributions of linear accelerations and rotational velocities were comparable for patients and age-matched healthy controls, comparison of power spectra revealed significant differences during more dynamic and challenging activities. Specifically, consistent with our prediction, head movement power spectra were significantly altered in patients during two tasks that required rapid online vestibular feedback: active repetitive jumping and walking on foam. Using computational methods, we analysed concurrently measured torso motion and identified increases in head-torso movement variability. Taken together, our results demonstrate that vestibular loss significantly alters head movement statistics and further suggest that increased variability and impaired feedback to internal models required for accurate motor control contribute to the observed changes.

Implicit motor learning within three trials

In motor learning, the slow development of implicit learning is traditionally taken for granted. While much is known about training performance during adaptation to a perturbation in reaches, saccades and locomotion, little is known about the time course of the underlying implicit processes during normal motor adaptation. Implicit learning is characterized by both changes in internal models and state estimates of limb position. Here, we measure both as reach aftereffects and shifts in hand localization in our participants, after every training trial. The observed implicit changes were near asymptote after only one to three perturbed training trials and were not predicted by a two-rate model's slow process that is supposed to capture implicit learning. Hence, we show that implicit learning is much faster than conventionally believed, which has implications for rehabilitation and skills training.

The Psychology of Reaching: Action Selection, Movement Implementation, and Sensorimotor Learning

The study of motor planning and learning in humans has undergone a dramatic transformation in the 20 years since this journal's last review of this topic. The behavioral analysis of movement, the foundational approach for psychology, has been complemented by ideas from control theory, computer science, statistics, and, most notably, neuroscience. The result of this interdisciplinary approach has been a focus on the computational level of analysis, leading to the development of mechanistic models at the psychological level to explain how humans plan, execute, and consolidate skilled reaching movements. This review emphasizes new perspectives on action selection and motor planning, research that stands in contrast to the previously dominant representation-based perspective of motor programming, as well as an emerging literature highlighting the convergent operation of multiple processes in sensorimotor learning.