The organization of quick corrections within a two-joint synergy in conditions of unexpected blocking and release of a fast movement

Motor synergy generalization framework for new targets in multi-planar and multi-directional reaching task

Humans can rapidly adapt to new situations, even though they have redundant degrees of freedom (d.f.). Previous studies in neuroscience revealed that human movements could be accounted for by low-dimensional control signals, known as motor synergies. Many studies have suggested that humans use the same repertories of motor synergies among similar tasks. However, it has not yet been confirmed whether the combinations of motor synergy repertories can be re-used for new targets in a systematic way. Here we show that the combination of motor synergies can be generalized to new targets that each repertory cannot handle. We use the multi-directional reaching task as an example. We first trained multiple policies with limited ranges of targets by reinforcement learning and extracted sets of motor synergies. Finally, we optimized the activation patterns of sets of motor synergies and demonstrated that combined motor synergy repertories were able to reach new targets that were not achieved with either original policies or single repertories of motor synergies. We believe this is the first study that has succeeded in motor synergy generalization for new targets in new planes, using a full 7-d.f. arm model, which is a realistic mechanical environment for general reaching tasks.

Role of the cortex in visuomotor control of arm stability

Whereas numerous motor control theories describe the control of arm trajectory during reach, the control of stabilization in a constant arm position (i.e., visuomotor control of arm posture) is less clear. Three potential mechanisms have been proposed for visuomotor control of arm posture: 1) increased impedance of the arm through co-contraction of antagonistic muscles, 2) corrective muscle activity via spinal/supraspinal reflex circuits, and/or 3) intermittent voluntary corrections to errors in position. We examined the cortical mechanisms of visuomotor control of arm posture and tested the hypothesis that cortical error networks contribute to arm stabilization. We collected electroencephalography (EEG) data from 10 young healthy participants across four experimental planar movement tasks. We examined brain activity associated with intermittent voluntary corrections of position error and antagonist co-contraction during stabilization. EEG beta-band (13-26 Hz) power fluctuations were used as indicators of brain activity, and coherence between EEG electrodes was used as a measure of functional connectivity between brain regions. Cortical activity in the sensory, motor, and visual areas during arm stabilization was similar to activity during volitional arm movements and was larger than activity during co-contraction of the arm. However, cortical connectivity between the sensorimotor and visual regions was higher during arm stabilization compared with volitional arm movements and co-contraction of the arm. The difference in cortical activity and connectivity between tasks might be attributed to an underlying visuomotor error network used to update motor commands for visuomotor control of arm posture.NEW & NOTEWORTHY We examined cortical activity and connectivity during control of stabilization in a constant arm position (i.e., visuomotor control of arm posture). Our findings provide evidence for cortical involvement during control of stabilization in a constant arm position. A visuomotor error network appears to be active and may update motor commands for visuomotor control of arm posture.

Shoulder reflexes integrate elbow information at "long-latency" delay throughout a corrective action

Previous studies have demonstrated a progression of function when healthy subjects counter a sudden mechanical load. Short-latency reflexes are linked to local stretch of the particular muscle and its antagonist. Long-latency reflexes integrate stretch information from both local sources and muscles crossing remote joints appropriate for a limb's mechanical interactions. Unresolved is how sensory information is processed throughout the corrective response, since capabilities at some time can be produced by circuits acting at that delay and at briefer delays. One possibility is that local abilities are always expressed at a short-latency delay and integrative abilities are always expressed at a long-latency delay. Alternatively, the neural circuits may be altered over time, leading to a temporal shift in expressing certain abilities; a refractory period could retard integrative responses to a second perturbation, whereas priming could enable integrative responses at short latency. We tested between these three hypotheses in a shoulder muscle by intermixing trials of step torque with either torque pulses ( experiment 1) or double steps of torque ( experiment 2). The second perturbation occurred at 35, 60, and 110 ms after the first perturbation to probe processing throughout the corrective action. The second perturbation reliably evoked short-latency responses in the shoulder muscle linked to only shoulder motion and long-latency responses linked to both shoulder and elbow motion. This pattern is best accounted by the continuous action of controllers with fixed functions. NEW & NOTEWORTHY Sudden displacement of the limb evokes a short-latency reflex, 20-50 ms, based on local muscle stretch and a long-latency reflex based on integrating muscle stretch at different joints. A novel double-perturbation paradigm tested if these abilities are temporally conserved throughout the corrective response or are shifted (retarded or delayed) due to functional changes in the responsible circuits. Multi-joint integration was reliably expressed at a long-latency delay consistent with the continuous operation of circuits with fixed abilities.

Long-latency reflexes for inter-effector coordination reflect a continuous state feedback controller

Successful performance in many everyday tasks requires compensating unexpected mechanical disturbance to our limbs and body. The long-latency reflex plays an important role in this process because it is the fastest response to integrate sensory information across several effectors, like when linking the elbow and shoulder or the arm and body. Despite the dozens of studies on inter-effector long-latency reflexes, there has not been a comprehensive treatment of how these reveal the basic control organization that sets constraints on any candidate model of neural feedback control such as optimal feedback control. We considered three contrasting ways that controllers can be organized: multiple independent controllers vs. a multiple-input multiple-output (MIMO) controller, a continuous feedback controller vs. an intermittent feedback controller, and a direct MIMO controller vs. a state feedback controller. Following a primer on control theory and review of the relevant evidence, we conclude that continuous state feedback control best describes the organization of inter-effector coordination by the long-latency reflex.

Generation of Human-Like Movement from Symbolized Information

An important function missing from current robotic systems is a human-like method for creating behavior from symbolized information. This function could be used to assess the extent to which robotic behavior is human-like because it distinguishes human motion from that of human-made machines created using currently available techniques. The purpose of this research is to clarify the mechanisms that generate automatic motor commands to achieve symbolized behavior. We design a controller with a learning method called tacit learning, which considers system–environment interactions, and a transfer method called mechanical resonance mode, which transfers the control signals into a mechanical resonance mode space (MRM-space). We conduct simulations and experiments that involve standing balance control against disturbances with a two-degree-of-freedom inverted pendulum and bipedal walking control with humanoid robots. In the simulations and experiments on standing balance control, the pendulum can become upright after a disturbance by adjusting a few signals in MRM-space with tacit learning. In the simulations and experiments on bipedal walking control, the robots realize a wide variety of walking by manually adjusting a few signals in MRM-space. The results show that transferring the signals to an appropriate control space is the key process for reducing the complexity of the signals from the environment and achieving diverse behavior.

Long-latency reflexes of elbow and shoulder muscles suggest reciprocal excitation of flexors, reciprocal excitation of extensors, and reciprocal inhibition between flexors and extensors

Postural corrections of the upper limb are required in tasks ranging from handling an umbrella in the changing wind to securing a wriggling baby. One complication in this process is the mechanical interaction between the different segments of the arm where torque applied at one joint induces motion at multiple joints. Previous studies have shown the long-latency reflexes of shoulder muscles (50-100 ms after a limb perturbation) account for these mechanical interactions by integrating information about motion of both the shoulder and elbow. It is less clear whether long-latency reflexes of elbow muscles exhibit a similar capability and what is the relation between the responses of shoulder and elbow muscles. The present study utilized joint-based loads tailored to the subjects' arm dynamics to induce well-controlled displacements of their shoulder and elbow. Our results demonstrate that the long-latency reflexes of shoulder and elbow muscles integrate motion from both joints: the shoulder and elbow flexors respond to extension at both joints, whereas the shoulder and elbow extensors respond to flexion at both joints. This general pattern accounts for the inherent flexion-extension coupling of the two joints arising from the arm's intersegmental dynamics and is consistent with spindle-based reciprocal excitation of shoulder and elbow flexors, reciprocal excitation of shoulder and elbow extensors, and across-joint inhibition between the flexors and extensors.

Long-latency reflexes account for limb biomechanics through several supraspinal pathways

Accurate control of body posture is enforced by a multitude of corrective actions operating over a range of time scales. The earliest correction is the short-latency reflex (SLR) which occurs between 20–45 ms following a sudden displacement of the limb and is generated entirely by spinal circuits. In contrast, voluntary reactions are generated by a highly distributed network but at a significantly longer delay after stimulus onset (greater than 100 ms). Between these two epochs is the long-latency reflex (LLR) (around 50–100 ms) which acts more rapidly than voluntary reactions but shares some supraspinal pathways and functional capabilities. In particular, the LLR accounts for the arm’s biomechanical properties rather than only responding to local muscle stretch like the SLR. This paper will review how the LLR accounts for the arm’s biomechanical properties and the supraspinal pathways supporting this ability. Relevant experimental paradigms include clinical studies, non-invasive brain stimulation, neural recordings in monkeys, and human behavioral studies. The sum of this effort indicates that primary motor cortex and reticular formation (RF) contribute to the LLR either by generating or scaling its structured response appropriate for the arm’s biomechanics whereas the cerebellum scales the magnitude of the feedback response. Additional putative pathways are discussed as well as potential research lines.

Fast feedback control involves two independent processes utilizing knowledge of limb dynamics

Corrective muscle responses occurring 50-100 ms after a mechanical perturbation are tailored to the mechanical features of the limb and its environment. For example, the evoked response by the shoulder's extensor muscle counters an imposed shoulder torque, rather than local shoulder motion, by integrating motion information from the shoulder and elbow appropriate for their dynamic interaction. Previous studies suggest that arm muscle activity within this epoch, alternately termed the R2/3 response, or long-latency reflex, reflects the summed result of two distinct components: an activity-dependent component which scales with the background muscle activity, and a task-dependent component which scales with the required vigor of the behavioral task. Here we examine how the knowledge of limb dynamics expressed during the shoulder muscle's R2/3 epoch is related to these two functional components. Subjects countered torque steps applied to their shoulder and/or elbow under different conditions of background torque and target size to recruit the activity-dependent and task-dependent component in varying degrees. Experiment 1 involved four torque perturbations, two levels of background torques and two target sizes; this design revealed that both background torque and target size impacted the shoulder's R2/3 activity, indicative of knowledge of limb dynamics. Experiment 2 involved two perturbation torques, five levels of background torque and two target sizes; this design demonstrated that the two factors had an independent impact on the R2/3 activity indicative of knowledge of limb dynamics. We conclude that a sophisticated feature of upper limb feedback control reflects the summation of two processes having a common capability.

Cerebellar damage diminishes long-latency responses to multijoint perturbations

Damage to the cerebellum can cause significant problems in the coordination of voluntary arm movements. One prominent idea is that incoordination stems from an inability to predictively account for the complex mechanical interactions between the arm's several joints. Motivated by growing evidence that corrective feedback control shares important capabilities and neural substrates with feedforward control, we asked whether cerebellar damage impacts feedback stabilization of the multijoint arm appropriate for the arm's intersegmental dynamics. Specifically, we tested whether cerebellar dysfunction impacts the ability of posterior deltoid to incorporate elbow motion in its long-latency response (R2 = 45-75 ms and R3 = 75-100 ms after perturbation) to an unexpected torque perturbation. Healthy and cerebellar-damaged subjects were exposed to a selected pattern of shoulder-elbow displacements to probe the response pattern from this shoulder extensor muscle. The healthy elderly subjects expressed a long-latency response linked to both shoulder and elbow motion, including an increase/decrease in shoulder extensor activity with elbow flexion/extension. Critically, cerebellar-damaged subjects displayed the normal pattern of activity in the R3 period indicating an intact ability to rapidly integrate multijoint motion appropriate to the arm's intersegmental dynamics. However, cerebellar-damaged subjects had a lower magnitude of activity that was specific to the long-latency period (both R2 and R3) and a slightly delayed onset of multijoint sensitivity. Taken together, our results suggest that the basic motor pattern of the long-latency response is housed outside the cerebellum and is scaled by processes within the cerebellum.

Maximization of learning speed in the motor cortex due to neuronal redundancy

Many redundancies play functional roles in motor control and motor learning. For example, kinematic and muscle redundancies contribute to stabilizing posture and impedance control, respectively. Another redundancy is the number of neurons themselves; there are overwhelmingly more neurons than muscles, and many combinations of neural activation can generate identical muscle activity. The functional roles of this neuronal redundancy remains unknown. Analysis of a redundant neural network model makes it possible to investigate these functional roles while varying the number of model neurons and holding constant the number of output units. Our analysis reveals that learning speed reaches its maximum value if and only if the model includes sufficient neuronal redundancy. This analytical result does not depend on whether the distribution of the preferred direction is uniform or a skewed bimodal, both of which have been reported in neurophysiological studies. Neuronal redundancy maximizes learning speed, even if the neural network model includes recurrent connections, a nonlinear activation function, or nonlinear muscle units. Furthermore, our results do not rely on the shape of the generalization function. The results of this study suggest that one of the functional roles of neuronal redundancy is to maximize learning speed.

There are overwhelmingly more neurons than muscles in the motor system. The functional roles of this neuronal redundancy remains unknown. Our analysis, which uses a redundant neural network model, reveals that learning speed reaches its maximum value if and only if the model includes sufficient neuronal redundancy. This result does not depend on whether the distribution of the preferred direction is uniform or a skewed bimodal, both of which have been reported in neurophysiological studies. We have confirmed that our results are consistent, regardless of whether the model includes recurrent connections, a nonlinear activation function, or nonlinear muscle units. Additionally, our results are the same when using either a broad or a narrow generalization function. These results suggest that one of the functional roles of neuronal redundancy is to maximize learning speed.

Motor abundance supports multitasking while standing

Many activities require simultaneous performance of multiple tasks. Motor redundancy may provide a key mechanism for multitasking, ensuring minimal inter-task interference. This study investigated the effect of performing two supra-postural tasks on postural stability. The component of joint configuration variance (JCV) reflecting flexible joint combinations (V(UCM)) that stabilize the center of mass (CoM) position and the component of JCV leading to variability (V(ORT)) of the CoM were determined using the Uncontrolled Manifold (UCM) approach. Subjects executed a targeting task alone or in combination with a ball-balancing task. UCM analysis revealed that the joints were coordinated such that their combined variance reflected primarily V(UCM), without a substantial effect on CoM position stability. Evidence for this flexible control strategy increased when the ball-balancing task was added to targeting, or when the index of difficulty of targeting increased, both without leading to substantial increases in V(ORT) or CoM position variance. The increase in joint variance when performing additional tasks without affecting adversely CoM position stability supports the hypothesis that the nervous system takes advantage of available motor redundancy for the successful performance of multiple tasks concurrently. Future work is needed to investigate the limits of this control scheme.

Heteronymous reflex connections in human upper limb muscles in response to stretch of forearm muscles

Torque motor produced stretch of upper limb muscles results in two distinct reflex peaks in the electromyographic activity. Whereas the short-latency reflex (SLR) response is mediated largely by the spinal monosynaptic reflex pathway, the longer-latency reflex (LLR) is suggested to involve a transcortical loop. For the SLRs, patterns of heteronymous monosynaptic Ia connections have been well-studied for a large number of muscles in the cat and in humans. For LLRs, information is available for perturbations to proximal joints, although the protocols for most of these studies did not focus on heteronymous connections. The main objective of the present study was to elicit both SLRs and LLRs in wrist flexors and extensors and to examine heteronymous connections from these muscles to elbow flexors (biceps brachii; BiBr) and extensors (triceps brachii; TriBr) and to selected distal muscles, including abductor pollicis longus (APL), first dorsal interosseous (FDI), abductor digiti minimi (ADM), and Thenars. The stretch of wrist flexors produced SLR and LLR peaks in APL, FDI, ADM, Thenars, and BiBr while simultaneously inducing inhibition of wrist extensors and TriBr. When wrist extensors were stretched, SLR and LLR peaks were observed in TriBr, whereas the primary wrist flexors, APL and BiBr, were inhibited; response patterns of FDI, ADM, and Thenars were less consistent. The main conclusions from the observed data are that: 1) as in the cat, afferents from wrist flexors and extensors make heteronymous connections with proximal and distal upper limb muscles; and 2) the strength of heteronymous connections is greater for LLRs than SLRs in the distal muscles, whereas the opposite is true for the proximal muscles. In the majority of observations, SLR and LLR excitatory peaks were observed together. However, on occasion, LLRs were observed without the SLR response in hand muscles when wrist extensors were stretched.

Long-Latency and Voluntary Responses to an Arm Displacement Can Be Rapidly Attenuated By Perturbation Offset

Feedback control of our limbs must account for the unexpected offset of mechanical perturbations. Here we examine the evoked activity of elbow flexor and extensor muscles to torque pulses lasting 22–152 ms and how torque offset impacts activity in the long-latency (45–100 ms) and voluntary epochs (120–180 ms). For each pulse width, we found a significant attenuation of muscle activity ∼30 ms after the offset of torque compared with when the torque was sustained. The brief time between the offset of torque and the attenuation of muscle activity implicates group I afferents acting through a spinal pathway, because this route is the only one fast enough and short enough to be responsible. Moreover, elbow muscle activity in the subsequent 20–45 ms following torque-offset was ∼35% smaller than when the torque was sustained. These results show that a fast spinal process can powerfully attenuate corrective responses of the arm to a torque perturbation.

Long-Latency Responses During Reaching Account for the Mechanical Interaction Between the Shoulder and Elbow Joints

Although considerable research indicates that reaching movements rely on knowledge of the arm's mechanical properties and environment to anticipate and counter predictable loads, far less research has examined whether this degree of sophistication is present for on-line corrections during reaching. Here we examine the R2/3 response to mechanical perturbations (45–100 ms, also called the long-latency reflex), which is highly flexible and includes the fastest possible contribution from primary motor cortex, a key neural substrate for self-initiated action. Torque perturbations were occasionally and unexpectedly applied to the subject's shoulder and/or elbow in the course of performing reaching movements. Critically, these perturbations would evoke different patterns of feedback corrections from a shoulder extensor muscle if it countered only the local shoulder displacement relative to unperturbed motion or accounted for the mechanical interactions between the shoulder and elbow joints and countered the underlying shoulder torque. Our results show that the earliest response (R1: 20–45 ms) reflected local shoulder displacement, whereas the R2/3 response (45–100 ms) reflected knowledge of multijoint dynamics. Moreover, the same pattern of feedback occurred whether the shoulder muscle helped initiate the movement (during its agonist phase) or helped terminate the movement (during its antagonist phase). These results contribute to the accumulating evidence that highly sophisticated feedback control underlies motor behavior and are consistent with a shared neural substrate, such as primary motor cortex, for feedforward and feedback control.

The valsalva maneuver revisited: the influence of voluntary breathing on isometric muscle strength

We assessed the effects of 4 voluntary breathing conditions on maximal voluntary isometric force of large muscle groups. Ten subjects performed maximum voluntary isometric contractions (MVICs) of knee flexion and extension, shoulder abduction and adduction, and elbow flexion and extension under all breathing conditions: normal breathing, forced inhalation, forced exhalation, and the Valsalva maneuver (VM). Forced exhalation significantly increased peak force during shoulder adduction, elbow extension, and knee extension MVIC tasks (p = 0.001, 0.024, and 0.002, respectively); the peak force during the Valsalva maneuver was not different from forced exhalation for all tested muscle groups. No voluntary breathing condition seemed to influence the peak force during the knee flexion, elbow flexion, and shoulder abduction MVIC tasks. The results demonstrate that voluntary breathing imposes a significant impact on isometric muscle strength. Given the increased cardiovascular risks associated with the Valsalva maneuver, it is highly recommended that forced exhalation be used during exercise at maximal levels, especially in repetitive repetitions.

Shared internal models for feedforward and feedback control

A child often learns to ride a bicycle in the driveway, free of unforeseen obstacles. Yet when she first rides in the street, we hope that if a car suddenly pulls out in front of her, she will combine her innate goal of avoiding an accident with her learned knowledge of the bicycle, and steer away or brake. In general, when we train to perform a new motor task, our learning is most robust if it updates the rules of online error correction to reflect the rules and goals of the new task. Here we provide direct evidence that, after a new feedforward motor adaptation, motor feedback responses to unanticipated errors become precisely task appropriate, even when such errors were never experienced during training. To study this ability, we asked how, if at all, do online responses to occasional, unanticipated force pulses during reaching arm movements change after adapting to altered arm dynamics? Specifically, do they change in a task-appropriate manner? In our task, subjects learned novel velocity-dependent dynamics. However, occasional force-pulse perturbations produced unanticipated changes in velocity. Therefore, after adaptation, task-appropriate responses to unanticipated pulses should compensate corresponding changes in velocity-dependent dynamics. We found that after adaptation, pulse responses precisely compensated these changes, although they were never trained to do so. These results provide evidence for a smart feedback controller which automatically produces responses specific to the learned dynamics of the current task. To accomplish this, the neural processes underlying feedback control must (1) be capable of accurate real-time state prediction for velocity via a forward model and (2) have access to recently learned changes in internal models of limb dynamics.

Neuromuscular and biomechanical factors codetermine the solution to motor redundancy in rhythmic multijoint arm movement

How the CNS deals with the issue of motor redundancy remains a central question for motor control research. Here we investigate the means by which neuromuscular and biomechanical factors interact to resolve motor redundancy in rhythmic multijoint arm movements. We used a two-df motorized robot arm to manipulate the dynamics of rhythmic flexion-extension (FE) and supination-pronation (SP) movements at the elbow-joint complex. Participants were required to produce rhythmic FE and SP movements, either in isolation, or in combination (at the phase relationship of their choice), while we recorded the activity of key bi-functional muscles. When performed in combination, most participants spontaneously produced an in-phase pattern of coordination in which flexion is synchronised with supination. The activity of the Biceps Brachii (BB), the strongest arm muscle which also has the largest moment arms in both flexion and supination was significantly higher for FE and SP performed in combination than in isolation, suggesting optimal exploitation of the mechanical advantage of this muscle. In a separate condition, participants were required to produce a rhythmic SP movement while a rhythmic FE movement was imposed by the motorized robot. Simulations based upon a musculoskeletal model of the arm demonstrated that in this context, the most efficient use of the force-velocity relationship of BB requires that an anti-phase pattern of coordination (flexion synchronized with pronation) be produced. In practice, the participants maintained the in-phase behavior, and BB activity was higher than for SP performed in isolation. This finding suggests that the neural organisation underlying the exploitation of bifunctional muscle properties, in the natural context, constrains the system to maintain the "natural" coordination pattern in an altered dynamic environment, even at the cost of reduced biomechanical efficiency. We suggest an important role for afference from the imposed movement in promoting the "natural" pattern. Practical implications for the emerging field of robot-assisted therapy and rehabilitation are briefly mentioned.

The Human Motor Control System's Response to Mechanical Perturbation: Should It, Can It and Does It Ensure Stability?

From among the diverse meanings of stability, the one the author adopts here is that the effects of a perturbation are opposed, and therefore small effects remain small. Except in linear systems, however, instability need not lead to unbounded motion and may actually be desirable when maneuverability is important. Moreover, properties of nerves, muscles, and tendons present serious challenges to stabilization. A review of observations from the motor control literature reveals that responses to perturbations in many common situations assist rather than resist the perturbation and are therefore presumably destabilizing. The observations encompass situations of position maintenance as well as impending or ongoing movement. The author proposes that the motor control system responds to a sudden perturbation by a pattern of muscle activity that mimics an accustomed voluntary movement, oblivious of stability considerations. What prevents runaway motion in the face of short-term instability appears to be voluntary intervention.