Search

The sense of agency is defined as the subjective experience that “I” am the one who is causing the action. Theoretical studies postulate that this subjective experience is developed through multistep processes extending from the sensorimotor to the cognitive level. However, it remains unclear how the brain processes such different levels of information and constitutes the neural substrates for the sense of agency. To answer this question, we combined two strategies: an experimental paradigm, in which self-agency gradually evolves according to sensorimotor experience, and a multivoxel pattern analysis. The combined strategies revealed that the sensorimotor, posterior parietal, anterior insula, and higher visual cortices contained information on self-other attribution during movement. In addition, we investigated whether the found regions showed a preference for self-other attribution or for sensorimotor information. As a result, the right supramarginal gyrus, a portion of the inferior parietal lobe (IPL), was found to be the most sensitive to self-other attribution among the found regions, while the bilateral precentral gyri and left IPL dominantly reflected sensorimotor information. Our results demonstrate that multiple brain regions are involved in the development of the sense of agency and that these show specific preferences for different levels of information.

Machine learning classifiers for psychiatric disorders using resting-state functional magnetic resonance imaging (rs-fMRI) have recently attracted attention as a method for directly examining relationships between neural circuits and psychiatric disorders. To develop accurate and generalizable classifiers, we compiled a large-scale, multi-site, multi-disorder neuroimaging database. The database comprises resting-state fMRI and structural images of the brain from 993 patients and 1,421 healthy individuals, as well as demographic information such as age, sex, and clinical rating scales. To harmonize the multi-site data, nine healthy participants (“traveling subjects”) visited the sites from which the above datasets were obtained and underwent neuroimaging with 12 scanners. All participants consented to having their data shared and analyzed at multiple medical and research institutions participating in the project, and 706 patients and 1,122 healthy individuals consented to having their data disclosed. Finally, we have published four datasets: 1) the SRPBS Multi-disorder Connectivity Dataset 2), the SRPBS Multi-disorder MRI Dataset (restricted), 3) the SRPBS Multi-disorder MRI Dataset (unrestricted), and 4) the SRPBS Traveling Subject MRI Dataset., Measurement(s)mental or behavioural disorder • brain measurement • Demographic DataTechnology Type(s)functional magnetic resonance imaging • magnetic resonance imaging • Resting State Functional Connectivity Magnetic Resonance ImagingFactor Type(s)age • sex • site • disorderSample Characteristic - OrganismHomo sapiens Machine-accessible metadata file describing the reported data: 10.6084/m9.figshare.14716329

Many studies have highlighted the difficulty inherent to the clinical application of fundamental neuroscience knowledge based on machine learning techniques. It is difficult to generalize machine learning brain markers to the data acquired from independent imaging sites, mainly due to large site differences in functional magnetic resonance imaging. We address the difficulty of finding a generalizable marker of major depressive disorder (MDD) that would distinguish patients from healthy controls based on resting-state functional connectivity patterns. For the discovery dataset with 713 participants from 4 imaging sites, we removed site differences using our recently developed harmonization method and developed a machine learning MDD classifier. The classifier achieved an approximately 70% generalization accuracy for an independent validation dataset with 521 participants from 5 different imaging sites. The successful generalization to a perfectly independent dataset acquired from multiple imaging sites is novel and ensures scientific reproducibility and clinical applicability., Biomarkers for psychiatric disorders based on neuroimaging data have yet to be put to practical use. This study overcomes the problems of inter-site differences in fMRI data by using a novel harmonization method, thereby successfully constructing a generalizable brain network marker of major depressive disorder across multiple imaging sites.

Adaptively recalibrating motor-sensory asynchrony is critical for animals to perceive self-produced action consequences. It is controversial whether motor- or sensory-related neural circuits recalibrate this asynchrony. By combining magnetoencephalography (MEG) and functional MRI (fMRI), we investigate the temporal changes in brain activities caused by repeated exposure to a 150-ms delay inserted between a button-press action and a subsequent flash. We found that readiness potentials significantly shift later in the motor system, especially in parietal regions (average: 219.9 ms), while visually evoked potentials significantly shift earlier in occipital regions (average: 49.7 ms) in the delay condition compared to the no-delay condition. Moreover, the shift in readiness potentials, but not in visually evoked potentials, was significantly correlated with the psychophysical measure of motor-sensory adaptation. These results suggest that although both motor and sensory processes contribute to the recalibration, the motor process plays the major role, given the magnitudes of shift and the correlation with the psychophysical measure.

When collecting large amounts of neuroimaging data associated with psychiatric disorders, images must be acquired from multiple sites because of the limited capacity of a single site. However, site differences represent a barrier when acquiring multisite neuroimaging data. We utilized a traveling-subject dataset in conjunction with a multisite, multidisorder dataset to demonstrate that site differences are composed of biological sampling bias and engineering measurement bias. The effects on resting-state functional MRI connectivity based on pairwise correlations because of both bias types were greater than or equal to psychiatric disorder differences. Furthermore, our findings indicated that each site can sample only from a subpopulation of participants. This result suggests that it is essential to collect large amounts of neuroimaging data from as many sites as possible to appropriately estimate the distribution of the grand population. Finally, we developed a novel harmonization method that removed only the measurement bias by using a traveling-subject dataset and achieved the reduction of the measurement bias by 29% and improvement of the signal-to-noise ratios by 40%. Our results provide fundamental knowledge regarding site effects, which is important for future research using multisite, multidisorder resting-state functional MRI data., This study was conducted under the contract research Grant Number JP18dm0307008, JP17dm0107044 (“Development of BMI Technologies for Clinical Application” of the Strategic Research Program for Brain Sciences), JP18dm0307002, JP18dm0307004, and JP18dm0307009 supported by the Japan Agency for Medical Research and Development (AMED). This study was also partially supported by the ImPACT Program of the Council for Science, Technology and Innovation (Cabinet Office, Government of Japan). H.I. was partially supported by JSPS KAKENHI 26120002. A.Y. was partially supported by JSPS KAKENHI 15J06788.

Although autism spectrum disorder (ASD) is a serious lifelong condition, its underlying neural mechanism remains unclear. Recently, neuroimaging-based classifiers for ASD and typically developed (TD) individuals were developed to identify the abnormality of functional connections (FCs). Due to over-fitting and interferential effects of varying measurement conditions and demographic distributions, no classifiers have been strictly validated for independent cohorts. Here we overcome these difficulties by developing a novel machine-learning algorithm that identifies a small number of FCs that separates ASD versus TD. The classifier achieves high accuracy for a Japanese discovery cohort and demonstrates a remarkable degree of generalization for two independent validation cohorts in the USA and Japan. The developed ASD classifier does not distinguish individuals with major depressive disorder and attention-deficit hyperactivity disorder from their controls but moderately distinguishes patients with schizophrenia from their controls. The results leave open the viable possibility of exploring neuroimaging-based dimensions quantifying the multiple-disorder spectrum., Autism spectrum disorder (ASD) is manifested by subtle but significant changes in the brain. Here, Yahata and colleagues devise a novel machine learning algorithm and develop a reliable ASD classifier based on brain functional connectivity, with which they quantitatively measure neuroimaging dimensions between ASD and other mental disorders.

Reaching toward a point target has been intensively studied in human motor control. However, little is known about reaching toward a redundant target, such as grasping a bar, in which the grasping point is irrelevant to the achievement of a task. We examined whether humans could solve the target-redundancy and control problems in a serial fashion or control their body without solving the target-redundancy problem. We equalized the target ranges between two reaching tasks: a point-to-point reaching task without target-redundancy and a point-to-bar reaching task with target-redundancy. In the both tasks, we measured hand viscoelasticity at movement end as parameters that reflect the adopted control strategy. As a result, the hand viscoelasticity in the point-to-bar reaching task was smaller than that in the point-to-point reaching task, even under the same kinematics. These results indicate that the hand viscoelasticity was modulated depending on the target-redundancy. Moreover, it is suggested that a human reaches toward a redundant target by effectively utilizing information of target redundancy rather than explicitly solving the target-redundancy problem.

Advances in functional magnetic resonance imaging have made it possible to provide real-time feedback on brain activity. Neurofeedback has been applied to therapeutic interventions for psychiatric disorders. Since many studies have shown that most psychiatric disorders exhibit abnormal brain networks, a novel experimental paradigm named connectivity neurofeedback, which can directly modulate a brain network, has emerged as a promising approach to treat psychiatric disorders. Here, we investigated the hypothesis that connectivity neurofeedback can induce the aimed direction of change in functional connectivity, and the differential change in cognitive performance according to the direction of change in connectivity. We selected the connectivity between the left primary motor cortex and the left lateral parietal cortex as the target. Subjects were divided into 2 groups, in which only the direction of change (an increase or a decrease in correlation) in the experimentally manipulated connectivity differed between the groups. As a result, subjects successfully induced the expected connectivity changes in either of the 2 directions. Furthermore, cognitive performance significantly and differentially changed from preneurofeedback to postneurofeedback training between the 2 groups. These findings indicate that connectivity neurofeedback can induce the aimed direction of change in connectivity and also a differential change in cognitive performance.

Perceived temporal interval between voluntary action and its outcome is shorter than that between involuntary action and its outcome (i.e., intentional binding). Although the effect is robust and extensively employed as a marker of sense of agency, the nature of intentional binding in multiple actions and outcomes remains unclear. We examined intentional binding in alternated action-outcome dyads. Participants actively or passively pressed a key, followed by a tone, and they again pressed the same key immediately after the preceding tone; resulting in four keypress-tone dyads in a trial. Participants reproduced the duration of alternated keypress-tone dyads or the temporal interval between a dyad embedded in the alternations. The reproduced duration was shorter in the active than in the passive condition, suggesting the intentional binding in action-outcome alternations. In contrast, the reproduced interval between a dyad was longer in the active condition and did not correlate with the reproduced duration of the alternations. These results suggest that subjective time during actions may rely not only on a general internal clock; rather, it may also be modulated by postdictive biases that are flexibly switched based on what we recall.

We grasp and dexterously manipulate an object through multi-1 digit synergy. In the framework of the uncontrolled manifold (UCM) hypothesis, multi-digit synergy is defined as the coordinated control mechanism of fingers to stabilize variable important for task success, e.g., total force. Previous studies reported anticipatory synergy adjustments (ASAs) that correspond to a drop of the synergy index before a quick change of the total force. The present study compared ASA’s properties with individual performances of feedforward force control to investigate a relationship of those. Subjects performed a total finger force production task that consisted of a phase in which subjects tracked target line with visual information and a phase in which subjects produced total force pulse without visual information. We quantified their multi-digit synergy through UCM analysis and observed significant ASAs before producing total force pulse. The time of the ASA initiation and the magnitude of the drop of the synergy index were significantly correlated with the error of force pulse, but not with the tracking error. Almost all subjects showed a significant increase of the variance that affected the total force. Our study directly showed that ASA reflects the individual performance of feedforward force control independently of target-tracking performance and suggests that the multi-digit synergy was weakened to adjust the multi-digit movements based on a prediction error so as to reduce the future error.

Timing is critical in determining the causal relationship between two events. Motor adaptation relies on the timing of actions and their results for determining which preceding control signals were responsible for subsequent error in the resulting movements. An artificially induced temporal delay in error feedback as short as 50 ms has been found to slow the learning rate of prism adaptation. Recent studies have demonstrated that our sense of simultaneity is flexibly adaptive when a persistent delay is inserted into visual feedback timing of one's own action. Therefore, judgments of "subjective simultaneity" (i.e. whether two events are simultaneous on a subjective basis) do not necessarily correspond to the actual simultaneity of physical events. We evaluated the effects of adaptation to a temporal shift of subjective simultaneity on prism adaptation by examining whether prism adaptation depends on physical timing or subjective timing. We found that after persistently experiencing an additional 100-ms delay in a pointing experiment, psychometric curves of the timing of judgments about the temporal order of touching and visual feedback were shifted by 40 ms, indicating that subjective simultaneity adapted. Next, while maintaining temporal adaptation, participants adapted to spatial displacement caused by a prism with and without an additional temporal delay in feedback. Learning speed was reliably predicted by physical timing but not by subjective timing. These results indicate that prism adaptation occurs independently of awareness of subjective timing and may be processed in primary motor areas that are thought to have fidelity with temporal relations.

Functional MRI was used to test predictions from a theory of the origin of human language. The gradual theory suggests that human language and tool-use skills have a similar hierarchical structure, and proposes that tool-manipulation skills are related to the origin and evolution of human language. Our results show an overlap of brain activity for perceiving language and using tools in Broca's area. The location of this overlap suggests that language and tool use share computational principles for processing complex hierarchical structures common to these two abilities. The involvement of monkeys' homologous region during tool use suggests that neural processes for computation of complex hierarchical structures exist in primates without language, and could have been exapted to support human grammatical ability.

Internal models are neural mechanisms that can mimic the input-output properties of controlled objects. Our studies have shown that: 1) an internal model for a novel tool is acquired in the cerebellum (Imamizu et al., 2000); 2) internal models are modularly organized in the cerebellum (Imamizu et al., 2003); 3) their outputs are sent to the premotor regions after learning (Tamada et al., 1999); and 4) the prefrontal and parietal regions contribute to the blending of the outputs (Imamizu et al., 2004). Here, we investigated changes in global neural networks resulting from the acquisition of a new internal model. Human subjects manipulated three types of rotating joystick whose cursor appeared at a position rotated 60 degrees, 110 degrees, or 160 degrees around the screen's center. In a pre-test after long-term training (5 days) for the 60 degrees and 160 degrees joysticks, brain activation was scanned during manipulation of the three joysticks. The subjects were then trained for the 110 degrees for only 25 min. In a post-test, activation was scanned using the same method as the pre-test. Comparisons of the post-test to the pre-test revealed that the volume of activation decreased in most of the regions where activation for the three rotations was observed. However, there was an increase in volume at a marginally significant level (p.08) only in the inferior-lateral cerebellum and only for the 110 degrees joystick. In the cerebral cortex, activation related to 110 degrees decreased in the prefrontal and parietal regions but increased in the premotor and supplementary motor area (SMA) regions. These results can be explained by a model in which outputs of the 60 degrees and 160 degrees internal models are blended by prefrontal and parietal regions to cope with the novel 110 degrees joystick before the 25-minute training; after the acquisition within the cerebellum of an internal model for the 110 degrees, output is directly sent to the premotor and SMA regions, and activation in these regions increases.

The purpose of this study is to identify the functional brain networks activated in relation to actual tool-use in humans. Although previous studies have identified brain activity related to tool-use gestures (Moll et al., 2000), they did not investigate the brain activity involved in such tool-use. We investigated brain activity using functional magnetic resonance imaging (fMRI) while human subjects mentally imagined using sixteen common tools and while they actually used them. Brain activity for both actual and imagined tool-use was found in the posterior part of the parietal cortex, in the supplementary motor area, and in the cerebellum. Under imagined tool-use conditions, we found brain activity in the premotor and right pars opercularis. Under actual tool-use conditions, we found it in the primary motor area, in the thalamus, and in the left pars opercularis. Our precise analysis in the cerebellum indicated that activity evoked by imagery was located significantly more lateral to that evoked by actual use. We found a relationship between activity in the tool imagery and execution conditions by comparing their t-value-weighted centroid of activation coordinates. Moreover, for half of the subjects the spatial distribution pattern for each tool was similar, suggesting that neural mechanisms contributing to skillful tool-use are modularly organized in the cerebellum.

UTokyo Research掲載「運動前の脳活動から反応の速さを予測することに成功」 URI: http://www.u-tokyo.ac.jp/ja/utokyo-research/research-news/prediction-of-reaction-speed-from-pre-movement-brain-activity.html, UTokyo Research "Prediction of reaction speed from pre-movement brain activity" URI: http://www.u-tokyo.ac.jp/en/utokyo-research/research-news/prediction-of-reaction-speed-from-pre-movement-brain-activity.html

SummarySkilled object manipulation requires knowledge, or internal models, of object dynamics relating applied force to motion [1–4], and our ability to handle myriad objects [5–6] indicates that the brain maintains multiple models [7–9]. Recent behavioral studies have shown that once learned, an internal model of an object with novel dynamics can be rapidly recruited and derecruited as the object is grasped and released [10–12]. We used event-related fMRI to investigate neural activity linked to grasping an object with recently learned dynamics in preparation for moving it after a delay. Subjects also performed two control tasks in which they either moved without the object in hand or applied isometric forces to the object. In all trials, subjects received a cue indicating which task to perform in response to a go signal delivered 5–10 s later. We examined BOLD responses during the interval between the cue and go and assessed the conjunction of the two contrasts formed by comparing the primary task to each control. The analysis revealed significant activity in the ipsilateral cerebellum and the contralateral and supplementary motor areas. We propose that these regions are involved in internal-model recruitment in preparation for movement execution.

Humans can acquire appropriate behaviors that maximize rewards on a trial-and-error basis. Recent electrophysiological and imaging studies have demonstrated that neural activity in the midbrain and ventral striatum encodes the error of reward prediction. However, it is yet to be examined whether the striatum is the main locus of reward-based behavioral learning. To address this, we conducted functional magnetic resonance imaging (fMRI) of a stochastic decision task involving monetary rewards, in which subjects had to learn behaviors involving different task difficulties that were controlled by probability. We performed a correlation analysis of fMRI data by using the explanatory variables derived from subject behaviors. We found that activity in the caudate nucleus was correlated with short-term reward and, furthermore, paralleled the magnitude of a subject's behavioral change during learning. In addition, we confirmed that this parallelism between learning and activity in the caudate nucleus is robustly maintained even when we vary task difficulty by controlling the probability. These findings suggest that the caudate nucleus is one of the main loci for reward-based behavioral learning.

Humans perform various motor tasks by coordinating the redundant motor elements in their bodies. The coordination of motor outputs is produced by motor commands, as well properties of the musculoskeletal system. The aim of this study was to dissociate the coordination of motor commands from motor outputs. First, we conducted simulation experiments where the total elbow torque was generated by a model of a simple human right and left elbow with redundant muscles. The results demonstrated that muscle tension with signal-dependent noise formed a coordinated structure of trial-to-trial variability of muscle tension. Therefore, the removal of signal-dependent noise effects was required to evaluate the coordination of motor commands. We proposed a method to evaluate the coordination of motor commands, which removed signal-dependent noise from the measured variability of muscle tension. We used uncontrolled manifold analysis to calculate a normalized index of synergy. Simulation experiments confirmed that the proposed method could appropriately represent the coordinated structure of the variability of motor commands. We also conducted experiments in which subjects performed the same task as in the simulation experiments. The normalized index of synergy revealed that the subjects coordinated their motor commands to achieve the task. Finally, the normalized index of synergy was applied to a motor learning task to determine the utility of the proposed method. We hypothesized that a large part of the change in the coordination of motor outputs through learning was because of changes in motor commands. In a motor learning task, subjects tracked a target trajectory of the total torque. The change in the coordination of muscle tension through learning was dominated by that of motor commands, which supported the hypothesis. We conclude that the normalized index of synergy can be used to evaluate the coordination of motor commands independently from the properties of the musculoskeletal system.

Quantitative examinations of internal representations for arm trajectory planning: minimum commanded torque change model. A number of invariant features of multijoint planar reaching movements have been observed in measured hand trajectories. These features include roughly straight hand paths and bell-shaped speed profiles where the trajectory curvatures between transverse and radial movements have been found to be different. For quantitative and statistical investigations, we obtained a large amount of trajectory data within a wide range of the workspace in the horizontal and sagittal planes (400 trajectories for each subject). A pair of movements within the horizontal and sagittal planes was set to be equivalent in the elbow and shoulder flexion/extension. The trajectory curvatures of the corresponding pair in these planes were almost the same. Moreover, these curvatures can be accurately reproduced with a linear regression from the summation of rotations in the elbow and shoulder joints. This means that trajectory curvatures systematically depend on the movement location and direction represented in the intrinsic body coordinates. We then examined the following four candidates as planning spaces and the four corresponding computational models for trajectory planning. The candidates were as follows: the minimum hand jerk model in an extrinsic-kinematic space, the minimum angle jerk model in an intrinsic-kinematic space, the minimum torque change model in an intrinsic-dynamic-mechanical space, and the minimum commanded torque change model in an intrinsic-dynamic-neural space. The minimum commanded torque change model, which is proposed here as a computable version of the minimum motor command change model, reproduced actual trajectories best for curvature, position, velocity, acceleration, and torque. The model’s prediction that the longer the duration of the movement the larger the trajectory curvature was also confirmed. Movements passing through via-points in the horizontal plane were also measured, and they converged to those predicted by the minimum commanded torque change model with training. Our results indicated that the brain may plan, and learn to plan, the optimal trajectory in the intrinsic coordinates considering arm and muscle dynamics and using representations for motor commands controlling muscle tensions.

The function of the lateral part of the human cerebellum was investigated through cerebro-cerebellar functional connectivity. We propose a laterality index method to reveal a functional and possibly anatomical pathway between the cerebral cortex and the cerebellum. The brain activity involved in learning a visually-guided tracking skill using a novel computer mouse was measured by functional magnetic resonance imaging. The imaging data analyzed using the method suggest that the simple lobule and semilunar lobule of the lateral cerebellum have connections with the pars opercularis and pars triangularis in the inferior frontal gyrus. A possible function of this cerebro-cerebellar communication loop is tool usage, which is in-between the cognitive and motor functions of the human cerebellum.

The elbow-joint angle and the shoulder-joint angle of participants aiming at targets were multiplied in an experiment that used a position-recording system and a cathode-ray tube screen. The linear transformation in joint angles (intrinsic coordinates) corresponded to a nonlinear transformation between the hand coordinates and the screen coordinates (extrinsic coordinates). We examined whether participants could learn this transformation in the intrinsic coordinates or in the extrinsic coordinates by investigating intermanual (between-hands) transfer under an intrinsically consistent condition and an extrinsically consistent condition. Positive intermanual transfer was observed in the former condition but not in the latter condition. Results suggest that participants can learn the linear transformation in joint angles under the intrinsic coordinates and that the central nervous system adaptively represents the intrinsic kinematics.

Behavioral studies have shown that humans can adapt to conflicting sensorimotor mappings that cause interference after intensive training. While previous research works indicate the involvement of distinct brain regions for different types of motor learning (e.g., kinematics vs dynamics), the neural mechanisms underlying joint adaptation to conflicting mappings within the same type of perturbation (e.g., different angles of visuomotor rotation) remain unclear. To reveal the neural substrates that represent multiple sensorimotor mappings, we examined whether different mappings could be classified with multivoxel activity patterns of functional magnetic resonance imaging data. Participants simultaneously adapted to opposite rotational perturbations (+90° and − 90°) during visuomotor tracking. To dissociate differences in movement kinematics with rotation types, we used two distinct patterns of target motion and tested generalization of the classifier between different combinations of rotation and motion types. Results showed that the rotation types were classified significantly above chance using activities in the primary sensorimotor cortex and the supplementary motor area, despite no significant difference in averaged signal amplitudes within the region. In contrast, low-level sensorimotor components, including tracking error and movement speed, were best classified using activities of the early visual cortex. Our results reveal that the sensorimotor cortex represents different visuomotor mappings, which permits joint learning and switching between conflicting sensorimotor skills.

Recent computational studies have proposed that the motor system acquires internal models of kinematic transformations, dynamic transformations, or both by learning. Computationally, internal models can be characterized by 2 extreme representations: structured and tabular (C. G. Atkeson, 1989). Tabular models do not need prior knowledge about the structure of the motor apparatus, but they lack the capability to generalize learned movements. Structured models, on the other hand, can generalize learned movements, but they require an analytical description of the motor apparatus. In investigating humans' capacity to generalize kinematic transformations, we examined which type of representation humans' motor system might use. Results suggest that internal representations are nonstructured and nontabular. Findings may be due to a neural network model with a medium number of neurons and synapses.

Reconstruction of movements from non-invasively recorded brain activity is a key technology for brain-machine interfaces (BMIs). However, electroencephalography (EEG) or magnetoencephalography (MEG) inevitably records a mixture of signals originating from many cortical regions, and thus it is not only less effective than invasive methods but also poses more difficulty for incorporating neuroscience knowledge. We combined two sparse Bayesian methods to overcome this difficulty. First, thousands of cortical currents were estimated on the order of millimeters and milliseconds by a hierarchical Bayesian MEG inverse method, and then a sparse regression method automatically selected only relevant cortical currents in accurate reconstruction of movements by a linear weighted sum of their time series. Using the combined methods, we reconstructed two-dimensional trajectories of the index fingertip during pointing movements to various directions by moving the wrist joint. A good generalization (reconstruction) performance was observed for test datasets: mean error between the predicted and actual positions was 15 mm, which was 7% of the path length of the required movement. The reconstruction accuracy of the proposed method was significantly higher than directly using MEG sensor signals. Moreover, spatial distribution and temporal characteristics of weight values revealed that the primary sensorimotor, higher motor, and parietal regions mainly contributed to the reconstruction with expected time courses. These results suggest that the combined sparse Bayesian methods provide effective means to predict movement trajectory from non-invasive brain activity directly related to sensorimotor control.

Switching of sensorimotor tasks may be classified into predictive switching based on contextual information and postdictive switching based on the error between sensorimotor feedback and predictions. We used functional neuroimaging to study the brain regions involved in each type of switching of internal models for visuomotor rotations (clockwise and counterclockwise rotations of visual feedback). The color of a cue presented before movement initiation corresponded to direction of rotation of the feedback in an instructed condition, but not in a noninstructed condition. Switching-related activity was identified as activity that transiently increased after the direction of rotation was changed. The switching-related activity in cue periods in the instructed condition, when a predictive switch is possible, was observed in the superior parietal lobule (SPL). However, the switching-related activity in feedback periods in the noninstructed condition, when prediction error is crucial for the postdictive switch, was observed in the inferior parietal lobule (IPL) and prefrontal cortex. The functional influence of the SPL on the lateral cerebellum, namely, a possible neural correlate for internal models, increased in the instructed condition, but the influence of the IPL on the cerebellum was increased in the noninstructed condition. We observed a rapid activity increase in the instructed condition and a gradual activity increase in the noninstructed condition mainly in the lateral occipito-temporal cortices (LOTC) and supplementary motor cortex (SMA). These results are consistent with separate mechanisms for predictive and postdictive switches and suggest that the LOTC and SMA receive output signals from appropriate internal models.

Recent computational and behavioral studies suggest that motor adaptation results from the update of multiple memories with different timescales. Here, we designed a model-based functional magnetic resonance imaging (fMRI) experiment in which subjects adapted to two opposing visuomotor rotations. A computational model of motor adaptation with multiple memories was fitted to the behavioral data to generate time-varying regressors of brain activity. We identified regional specificity to timescales: in particular, the activity in the inferior parietal region and in the anterior-medial cerebellum was associated with memories for intermediate and long timescales, respectively. A sparse singular value decomposition analysis of variability in specificities to timescales over the brain identified four components, two fast, one middle, and one slow, each associated with different brain networks. Finally, a multivariate decoding analysis showed that activity patterns in the anterior-medial cerebellum progressively represented the two rotations. Our results support the existence of brain regions associated with multiple timescales in adaptation and a role of the cerebellum in storing multiple internal models.

Previous functional imaging experiments in humans showed activation increases in the posterior superior temporal gyrus and sulcus during observation of geometrical shapes whose movements appear intentional or goal-directed. We modeled a chase scenario between two objects, in which the chasing object used different strategies to reach the target object: The chaser either followed the target's path or appeared to predict its end position. Activation in the superior temporal gyrus of human observers was greater when the chaser adopted a predict rather than a follow strategy. Attending to the chaser's strategy induced slightly greater activation in the left superior temporal gyrus than attending to the outcome of the chase. These data implicate the superior temporal gyrus in the identification of objects displaying complex goal-directed motion.

An internal model is a neural mechanism that can mimic the input–output properties of a controlled object such as a tool. Recent research interests have moved on to how multiple internal models are learned and switched under a given context of behavior. Two representative computational models for task switching propose distinct neural mechanisms, thus predicting different brain activity patterns in the switching of internal models. In one model, called the mixture-of-experts architecture, switching is commanded by a single executive called a “gating network,” which is different from the internal models. In the other model, called the MOSAIC (MOdularSelectionAndIdentification forControl), the internal models themselves play crucial roles in switching. Consequently, the mixture-of-experts model predicts that neural activities related to switching and internal models can be temporally and spatially segregated, whereas the MOSAIC model predicts that they are closely intermingled. Here, we directly examined the two predictions by analyzing functional magnetic resonance imaging activities during the switching of one common tool (an ordinary computer mouse) and two novel tools: a rotated mouse, the cursor of which appears in a rotated position, and a velocity mouse, the cursor velocity of which is proportional to the mouse position. The switching and internal model activities temporally and spatially overlapped each other in the cerebellum and in the parietal cortex, whereas the overlap was very small in the frontal cortex. These results suggest that switching mechanisms in the frontal cortex can be explained by the mixture-of-experts architecture, whereas those in the cerebellum and the parietal cortex are explained by the MOSAIC model.

Human capabilities in dexterously manipulating many different tools suggest modular neural organization at functional levels, but anatomical modularity underlying the capabilities has yet to be demonstrated. Although modularity in phylogenetically older parts of the cerebellum is well known, comparable modularity in the lateral cerebellum for cognitive functions remains unknown. We investigated these issues by functional MRI (fMRI) based on our previous findings of a cerebellar internal model of a tool. After subjects intensively learned to manipulate two novel tools (the rotated mouse whose cursor appeared at a rotated position, and the velocity mouse whose cursor velocity was proportional to the mouse position), they could easily switch between the two. The lateral and posterior cerebellar activities for the two different tools were spatially segregated, and their overlaps were

Internal models are neural mechanisms that can mimic the input-output or output-input properties of the motor apparatus and external objects. Forward internal models predict sensory consequences from efference copies of motor commands. There is growing acceptance of the idea that forward models are important in sensorimotor integration as well as in higher cognitive function, but their anatomical loci and neural mechanisms are still largely unknown. Some of the most convincing evidence that the central nervous system (CNS) makes use of forward models in sensory motor control comes from studies on grip force-load force coupling. We first present a brief review of recent computational and behavioral studies that provide decisive evidence for the utilization of forward models in grip force-load force coupling tasks. Then, we used functional magnetic resonance imaging (fMRI) to measure the brain activity related to this coupling and demonstrate that the cerebellum is the most likely site for forward models to be stored.

The cerebellum is thought to help coordinate movement. We tested this using functional magnetic resonance imaging (fMRI) of the human brain during visually guided tracking tasks requiring varying degrees of eye-hand coordination. The cerebellum was more active during independent rather than coordinated eye and hand tracking. However, in three further tasks, we also found parametric increases in cerebellar blood oxygenation signal (BOLD) as eye-hand coordination increased. Thus, the cerebellar BOLD signal has a non-monotonic relationship to tracking performance, with high activity during both coordinated and independent conditions. These data provide the most direct evidence from functional imaging that the cerebellum supports motor coordination. Its activity is consistent with roles in coordinating and learning to coordinate eye and hand movement.

Theories of motor control postulate that the brain uses internal models of the body to control movements accurately. Internal models are neural representations of how, for instance, the arm would respond to a neural command, given its current position and velocity. Previous studies have shown that the cerebellar cortex can acquire internal models through motor learning. Because the human cerebellum is involved in higher cognitive function as well as in motor control, we propose a coherent computational theory in which the phylogenetically newer part of the cerebellum similarly acquires internal models of objects in the external world. While human subjects learned to use a new tool (a computer mouse with a novel rotational transformation), cerebellar activity was measured by functional magnetic resonance imaging. As predicted by our theory, two types of activity were observed. One was spread over wide areas of the cerebellum and was precisely proportional to the error signal that guides the acquisition of internal models during learning. The other was confined to the area near the posterior superior fissure and remained even after learning, when the error levels had been equalized, thus probably reflecting an acquired internal model of the new tool.

The learning process of reaching movements was examined under novel environments whose kinematic and dynamic properties were altered. We used a kinematic transformation (visuomotor rotation), a dynamic transformation (viscous curl field), and a combination of these transformations. When the subjects learned the combined transformation, reaching errors were smaller if the subject first learned the separate kinematic and dynamic transformations. Reaching errors under the kinematic (but not the dynamic) transformation were smaller if subjects first learned the combined transformation. These results suggest that the brain learns multiple internal models to compensate for each transformation and has some ability to combine and decompose these internal models as called for by the occasion.

To assess the functional locus of visual-motor learning, the computational concepts of "task level" programming (determination of the trajectory of a hand during arm reaching in the Cartesian coordinates) and "manipulator level" programming (determination of the joint coordinates) was adopted. Because the former is likely to be hand nonspecific and the latter is hand specific, it is assumed that learning at the task level should be transferred to the unpracticed hand, whereas that at the manipulator level it should not. Under this assumption, the paradigm of intermanual transfer was used in an aiming task under rotated visual feedback. Nearly 100% intermanual transfer from the practiced hand to the unpracticed hand in the performance time of aiming was found, concluding that the locus of visual-motor learning should be at the task level rather than at the manipulator level.

Many evidences suggest that the central nervous system (CNS) acquires and switches internal models for adaptive control in various environments. However, little is known about the neural mechanisms responsible for the switching. A recent computational model for simultaneous learning and switching of internal models proposes two separate switching mechanisms: a predictive mechanism purely based on contextual information and a postdictive mechanism based on the difference between actual and predicted sensorimotor feedbacks. This model can switch internal models solely based on contextual information in a predictive fashion immediately after alteration of the environment. Here we show that when subjects simultaneously adapted to alternating blocks of opposing visuomotor rotations, explicit contextual information about the rotations improved the initial performance at block alternations and asymptotic levels of performance within each block but not readaptation speeds. Our simulations using separate switching mechanisms duplicated these effects of contextual information on subject performance and suggest that improvement of initial performance was caused by improved accuracy of the predictive switch while adaptation speed corresponds to a switch dependent on sensorimotor feedback. Simulations also suggested that a slow change in output signals from the switching mechanisms causes contamination of motor commands from an internal model used in the previous context (anterograde interference) and partial destruction of internal models (retrograde interference). Explicit contextual information prevents destruction and assists memory retention by improving the changes in output signals. Thus, the asymptotic levels of performance improved.

Real-time acquisition of electromyography (EMG) during functional Magnetic Resonance Imaging (fMRI) provides a novel method of controlling motor experiments in the scanner using feedback of EMG. Due to the redundancy in the human muscle system this is not possible from recordings of joint torque and kinematics alone, as these provide no information about individual muscle activation. This is particularly critical during brain imaging as brain activations are not only related to joint torques and kinematics but are also related to individual muscle activation. However, EMG collected during imaging is corrupted by large artifacts induced by the varying magnetic fields and radio-frequency (RF) pulses in the scanner. Methods proposed in literature for artifact removal are complex, computationally expensive and difficult to implement for real-time noise removal. We describe an acquisition system and algorithm which enables real-time acquisition for the first time. The algorithm removes particular frequencies from the EMG spectrum in which the noise is concentrated. Although this decreases the power content of the EMG, this method provides excellent estimates of EMG with good resolution. Comparisons show that the cleaned EMG obtained with the algorithm is, like actual EMG, very well correlated with joint torque and can thus be used for real-time visual feedback during functional studies.