Proprioceptive population coding of limb position in humans

Contrasting action and posture coding with hierarchical deep neural network models of proprioception

Biological motor control is versatile, efficient, and depends on proprioceptive feedback. Muscles are flexible and undergo continuous changes, requiring distributed adaptive control mechanisms that continuously account for the body's state. The canonical role of proprioception is representing the body state. We hypothesize that the proprioceptive system could also be critical for high-level tasks such as action recognition. To test this theory, we pursued a task-driven modeling approach, which allowed us to isolate the study of proprioception. We generated a large synthetic dataset of human arm trajectories tracing characters of the Latin alphabet in 3D space, together with muscle activities obtained from a musculoskeletal model and model-based muscle spindle activity. Next, we compared two classes of tasks: trajectory decoding and action recognition, which allowed us to train hierarchical models to decode either the position and velocity of the end-effector of one's posture or the character (action) identity from the spindle firing patterns. We found that artificial neural networks could robustly solve both tasks, and the networks' units show tuning properties similar to neurons in the primate somatosensory cortex and the brainstem. Remarkably, we found uniformly distributed directional selective units only with the action-recognition-trained models and not the trajectory-decoding-trained models. This suggests that proprioceptive encoding is additionally associated with higher-level functions such as action recognition and therefore provides new, experimentally testable hypotheses of how proprioception aids in adaptive motor control.

Passive Proprioceptive Training Alters the Sensitivity of Muscle Spindles to Imposed Movements

Humans rely on precise proprioceptive feedback from our muscles, which is important in both the acquisition and execution of movements, to perform daily activities. Somatosensory input from the body shapes motor learning through central processes, as demonstrated for tasks using the arm, under active (self-generated) and passive conditions. Presently, we investigated whether passive movement training of the ankle increased proprioceptive acuity (psychophysical experiment) and whether it changed the peripheral proprioceptive afferent signal (microneurography experiment). In the psychophysical experiment, the ankle of 32 healthy human participants was moved passively using pairs of ramp-and-hold movements in different directions. In a pretraining test, participants made judgements about the movement direction in a two-alternative forced choice paradigm. Participants then underwent passive movement training, but only half were cued for learning, where a reference position was signaled by a sound and the participant had to learn to recognize this position; they then completed a post-training test. In a paradigm using the same setup, nine healthy participants underwent microneurography recordings of Ia muscle afferents from the peroneal nerve, where all were cued during training. In the psychophysical experiment, proprioceptive acuity improved with training only in the cued group. In the microneurography experiment, we found that muscle afferent firing was modulated, via an increase in the dynamic index, after training. We suggest that changes in muscle afferent input from the periphery can contribute to and support central perceptual and motor learning, as shown under passive conditions using ankle movements, which may be exploited for movement rehabilitation.

Relative posture between head and finger determines perceived tactile direction of motion

The hand explores the environment for obtaining tactile information that can be fruitfully integrated with other functions, such as vision, audition, and movement. In theory, somatosensory signals gathered by the hand are accurately mapped in the world-centered (allocentric) reference frame such that the multi-modal information signals, whether visual-tactile or motor-tactile, are perfectly aligned. However, an accumulating body of evidence indicates that the perceived tactile orientation or direction is inaccurate; yielding a surprisingly large perceptual bias. To investigate such perceptual bias, this study presented tactile motion stimuli to healthy adult participants in a variety of finger and head postures, and requested the participants to report the perceived direction of motion mapped on a video screen placed on the frontoparallel plane in front of the eyes. Experimental results showed that the perceptual bias could be divided into systematic and nonsystematic biases. Systematic bias, defined as the mean difference between the perceived and veridical directions, correlated linearly with the relative posture between the finger and the head. By contrast, nonsystematic bias, defined as minor difference in bias for different stimulus directions, was highly individualized, phase-locked to stimulus orientation presented on the skin. Overall, the present findings on systematic bias indicate that the transformation bias among the reference frames is dominated by the finger-to-head posture. Moreover, the highly individualized nature of nonsystematic bias reflects how information is obtained by the orientation-selective units in the S1 cortex.

A new method to elicit and measure movement illusions in stroke by means of muscle tendon vibration: the Standardized Kinesthetic Illusion Procedure (SKIP)

Purpose: Muscle tendon vibration (MTV) strongly activates muscle spindles and can evoke kinaesthetic illusions. Although potentially relevant for sensorimotor rehabilitation in stroke, MTV is scarcely used in clinical practice, likely because of the absence of standardised procedures to elicit and characterise movement illusions. This work developed and validated a Standardised Kinaesthetic Illusion Procedure (SKIP) to favour the use of MTV-induced illusions in clinical settings.Materials and methods: SKIP scores were obtained in 15 individuals with chronic stroke and 18 age- and gender-matched healthy counterparts. A further 13 healthy subjects were tested to provide more data with the general population. MTV was applied over the Achilles tendon and SKIP scoring system characterised the clearness and direction of the illusions of ankle dorsiflexion movements.Results: All healthy and stroke participants perceived movement illusions. SKIP scores on the paretic side were significantly lower compared to the non paretic and healthy. Illusions were less clear and sometimes in unexpected directions with the impaired ankle, but still possible to elicit in the presence of sensorimotor deficits.Conclusions: SKIP represents an ancillary and potentially useful clinical method to elicit and characterise illusions of movements induced by MTV. SKIP could be relevant to further assess the processing of proprioceptive afferents in stroke and their potential impact on motor control and recovery. It may be used to guide therapy and improve sensorimotor recovery. Future work is needed to investigate the metrological properties of our method (reliability, responsiveness, etc.), and also the neurophysiological underpinnings of MTV-induced illusions.

Linear Integration of Tactile and Non-tactile Inputs Mediates Estimation of Fingertip Relative Position

While skin, joints and muscles receptors alone provide lower level information about individual variables (e.g., exerted limb force and limb displacement), the distance between limb endpoints (i.e., relative position) has to be extracted from high level integration of somatosensory and motor signals. In particular, estimation of fingertip relative position likely involves more complex sensorimotor transformations than those underlying hand or arm position sense: the brain has to estimate where each fingertip is relative to the hand and where fingertips are relative to each other. It has been demonstrated that during grasping, feedback of digit position drives rapid adjustments of fingers force control. However, it has been shown that estimation of fingertips' relative position can be biased by digit forces. These findings raise the question of how the brain combines concurrent tactile (i.e., cutaneous mechanoreceptors afferents induced by skin pressure and stretch) and non-tactile (i.e., both descending motor command and joint/muscle receptors signals associated to muscle contraction) digit force-related inputs for fingertip distance estimation. Here we addressed this question by quantifying the contribution of tactile and non-tactile force-related inputs for the estimation of fingertip relative position. We asked subjects to match fingertip vertical distance relying only on either tactile or non-tactile inputs from the thumb and index fingertip, and compared their performance with the condition where both types of inputs were combined. We found that (a) the bias in the estimation of fingertip distance persisted when tactile inputs and non-tactile force-related signals were presented in isolation; (b) tactile signals contributed the most to the estimation of fingertip distance; (c) linear summation of the matching errors relying only on either tactile or non-tactile inputs was comparable to the matching error when both inputs were simultaneously available. These findings reveal a greater role of tactile signals for sensing fingertip distance and suggest a linear integration mechanism with non-tactile inputs for the estimation of fingertip relative position.

Firing properties of muscle spindles supplying the intrinsic foot muscles of humans in unloaded and freestanding conditions

We recently developed an approach for recording from muscle spindles in the intrinsic muscles of the foot in freestanding humans by inserting a tungsten microelectrode into the posterior tibial nerve behind the medial malleolus of the ankle. Here we characterize the behavior of muscle spindles in the small muscles of the foot in 1) seated subjects with the leg horizontal and the foot naturally plantarflexed and 2) standing subjects. In the first study, recordings were made from 26 muscle spindle afferents located within flexor digiti minimi brevis ( n = 4), abductor digiti minimi ( n = 3), quadratus plantae ( n = 3), plantar interossei ( n = 4), flexor digitorum brevis ( n = 3), dorsal interossei ( n = 2), and lumbricals ( n = 2), with one each supplying abductor hallucis, adductor hallucis, and flexor hallucis brevis. The identity of another two muscle afferents was unknown. The majority of the units were silent at rest, only seven (27%) being spontaneously active. Because of the anatomic constraints of the foot, some spindles supplying muscles acting on the toes responded to movements of one or more digits. In the second study, 12 muscle spindle afferents were examined during standing. The ongoing discharge of eight spindle afferents covaried with changes in the center of pressure during postural sway. We conclude that the majority of spindle endings in the small muscles of the foot are silent at rest, which may allow them to encode changes in conformation of the foot when it is loaded during standing. Moreover, these muscle spindle afferents can provide useful proprioceptive information during standing and postural sway. NEW & NOTEWORTHY We have characterized the firing properties of muscle spindles in the intrinsic muscles of the human foot for the first time. The majority of the spindle endings are silent in seated subjects, and most fire tonically during standing, their discharge covarying with center of pressure during postural sway. We conclude that spindle endings in the intrinsic muscles of the foot provide useful proprioceptive information during free standing.

Functional properties of human muscle spindles

Muscle spindles are ubiquitous encapsulated mechanoreceptors found in most mammalian muscles. There are two types of endings, primary and secondary, and both are sensitive to changes in muscle length and velocity, with the primary endings having a greater dynamic sensitivity. Unlike other mechanoreceptors in the somatosensory system, muscle spindles are unique in possessing motor innervation, via γ-motoneurons (fusimotor neurons), that control their sensitivity to stretch. Much of what we know about human muscles spindles comes from studying the behavior of their afferents via intraneural microelectrodes (microneurography) inserted into accessible peripheral nerves. We review the functional properties of human muscle spindles, comparing and contrasting with what we know about the functions of muscle spindles studied in experimental animals. As in the cat, many human muscle spindles possess a background discharge that is related to the degree of muscle stretch, but mean firing rates are much lower (~10 Hz). They can faithfully encode changes in muscle fascicle length in passive conditions, but higher level extraction of information is required by the central nervous system to measure changes in muscle length during muscle contraction. Moreover, although there is some evidence supporting independent control of human muscle spindles via fusimotor neurons, any effects are modest compared with the clearly independent control of fusimotor neurons observed in the cat.

Precise coding of ankle angle and velocity by human calf muscle spindles

Human standing balance control requires the integration of sensory feedback to produce anticipatory, stabilizing ankle torques. However, the ability of human triceps surae muscle spindles to provide reliable sensory feedback regarding the small, slow ankle movements that occur during upright standing has recently come under question. We performed microneurography to directly record axon potentials from single muscle spindle afferents in the human triceps surae during servo-controlled movement of the ankle joint. To simulate movements of the ankle while standing, we delivered random 90-s dorsiflexion/plantar flexion oscillations of the ankle joint, with a peak-to-peak amplitude of 0.7° and frequency content below 0.5Hz. In roughly half of the trials (46%), participants held a low-level, near-isometric contraction of the triceps surae muscles. We demonstrate that afferent activity in a population of muscle spindles closely reflects ankle movements at frequencies and amplitudes characteristic of human standing. Four out of five soleus spindles, and three out of seven gastrocnemius spindles coded for at least a single frequency component of anteroposterior ankle rotation. Concatenating within muscles, coherence was significantly greater for soleus spindles at all stimulus frequencies. Voluntary contraction of the parent muscle reduced spindle sensitivity, but only significantly near the mean power frequency of the stimulus (∼0.3Hz). In conclusion, these results provide direct evidence that triceps surae muscle spindles are potentially capable of providing important sensory feedback for the control of human standing balance.

Shoulder elevation affects joint position sense and muscle activation differently in upright and supine body orientations

OBJECTIVE

Investigate the effects of shoulder elevation on repositioning errors in upright and supine body orientations, and examine these effects on anterior and posterior deltoid muscle activation. We hypothesized decreased errors, and altered anterior and posterior deltoid activation with increasing elevation, in both orientations.

DESIGN

Crossover trial.

SETTING

University laboratory.

PARTICIPANTS

Thirty-five college-aged participants.

INTERVENTION

Subjects attempted to replicate target positions of various elevation angles in upright and supine body orientations. Also, anterior and posterior deltoid activation was recorded in each shoulder position and body orientation.

MAIN OUTCOME MEASURES

Vector and variable repositioning errors, anterior and posterior deltoid percentage of maximal contraction.

RESULTS

Vector error was greater in supine compared to upright at 90° and 110°, but not at 70°. Variable error was larger in supine than upright, but was unaffected by elevation. Anterior deltoid activation increased with elevation in the upright posture only. Posterior deltoid activation increased with elevation across postures.

CONCLUSIONS

Muscle activation, external torque, and cutaneous sensations may combine to provide afferent feedback, and be used with centrally-generated signals to interpret the state of the limb during movement. Clinicians may prescribe open kinetic chain exercises in the upright posture with the shoulder elevated approximately 90-100°.

An Assessment of Six Muscle Spindle Models for Predicting Sensory Information during Human Wrist Movements

Background: The muscle spindle is an important sensory organ for proprioceptive information, yet there have been few attempts to use Shannon information theory to quantify the capacity of human muscle spindles to encode sensory input.

Methods: Computer simulations linked kinematics, to biomechanics, to six muscle spindle models that generated predictions of firing rate. The predicted firing rates were compared to firing rates of human muscle spindles recorded during a step-tracking (center-out) task to validate their use. The models were then used to predict firing rates during random movements with statistical properties matched to the ergonomics of human wrist movements. The data were analyzed for entropy and mutual information.

Results: Three of the six models produced predictions that approximated the firing rate of human spindles during the step-tracking task. For simulated random movements these models predicted mean rates of 16.0 ± 4.1 imp/s (mean ± SD), peak firing rates <50 imp/s and zero firing rate during an average of 25% of the movement. The average entropy of the neural response was 4.1 ± 0.3 bits and is an estimate of the maximum information that could be carried by muscles spindles during ecologically valid movements. The information about tendon displacement preserved in the neural response was 0.10 ± 0.05 bits per symbol; whereas 1.25 ± 0.30 bits per symbol of velocity input were preserved in the neural response of the spindle models.

Conclusions: Muscle spindle models, originally based on cat experiments, have predictive value for modeling responses of human muscle spindles with minimal parameter optimization. These models predict more than 10-fold more velocity over length information encoding during ecologically valid movements. These results establish theoretical parameters for developing neuroprostheses for proprioceptive function.

Three tools for the real-time simulation of embodied spiking neural networks using GPUs

This paper presents a toolbox of solutions that enable the user to construct biologically-inspired spiking neural networks with tens of thousands of neurons and millions of connections that can be simulated in real time, visualized in 3D and connected to robots and other devices. NeMo is a high performance simulator that works with a variety of neural and oscillator models and performs parallel simulations on either GPUs or multi-core processors. SpikeStream is a visualization and analysis environment that works with NeMo and can construct networks, store them in a database and visualize their activity in 3D. The iSpike library provides biologically-inspired conversion between real data and spike representations to support work with robots, such as the iCub. Each of the tools described in this paper can be used independently with other software, and they also work well together.

Flexible kinesthetic distance perception: when do your arms tell you how far you have walked?

Given the flexible organization of locomotion evidenced in the many ways the limbs can be coordinated, the authors explored the potentially correspondingly flexible organization of nonvisual (kinesthetic) distance perception. As kinesthetic distance perception is known to be affected by how the limbs are coordinated, the authors probed the potential perceptual contribution of the arms during locomotion by manipulating arm-leg coordination patterns in blind-walked distance-matching tasks. Whereas manipulation of arm-leg coordination for walking with free-swinging arms had no observable perceptual consequences, comparable manipulation for walking with hiking poles did affect distance matching. These results suggest that under conditions in which the arms act to propel the body (e.g., crawling or stair-climbing) a person's nonvisual sense of movement is conveyed in the coordinated actions of all four limbs.

Temporal features of human tendon vibration illusions

Muscle spindles provide information about the position and movement of our bodies. One method for investigating spindle signals is tendon vibration. Vibration of flexor tendons can produce illusions of extension, and vibration of extensor tendons can produce illusions of flexion. Here we estimate the temporal resolution and persistence of these illusions. In Experiments 1 and 2, sequences of alternating vibration of wrist flexor and extensor tendons produced position illusions that varied with alternation period. When vibrations alternated at 1 Hz or slower, perceived position at the end of the sequence depended on the last vibration. When vibrations alternated every 0.3 s, perceived position was independent of the last vibration. Experiment 2 verified and extended these results using more trials and concurrent electromyographic recording. Although tendon vibrations sometimes induce reflexive muscle activity, we found no evidence that such activity contributed to these effects. Experiment 3 investigated how long position sense is retained when not updated by current information from spindles. Our first experiments suggested that vibrating antagonistic tendons simultaneously could produce conflicting inputs, leaving position sense reliant on memory of position prior to vibration onset. We compared variability in position sense after different durations of such double vibration. After 12 s of double vibration, variability across trials exceeded levels predicted from vibrations of flexor or extensor tendons alone. This suggests that position sense memory had decayed too much to substitute for the current conflicting sensory information. Together, our results provide novel, quantitative insight into the temporal properties of tendon vibration illusions.

Functional changes in muscle afferent neurones in an osteoarthritis model: implications for impaired proprioceptive performance

Background

Impaired proprioceptive performance is a significant clinical issue for many who suffer osteoarthritis (OA) and is a risk factor for falls and other liabilities. This study was designed to evaluate weight-bearing distribution in a rat model of OA and to determine whether changes also occur in muscle afferent neurones.

Methodology/Principal Findings

Intracellular recordings were made in functionally identified dorsal root ganglion neurones in acute electrophysiological experiments on the anaesthetized animal following measurements of hind limb weight bearing in the incapacitance test. OA rats but not naïve control rats stood with less weight on the ipsilateral hind leg (P = 0.02). In the acute electrophysiological experiments that followed weight bearing measurements, action potentials (AP) elicited by electrical stimulation of the dorsal roots differed in OA rats, including longer AP duration (P = 0.006), slower rise time (P = 0.001) and slower maximum rising rate (P = 0.03). Depolarizing intracellular current injection elicited more APs in models than in naïve muscle afferent neurones (P = 0.01) indicating greater excitability. Axonal conduction velocity in model animals was slower (P = 0.04).

Conclusions/Significance

The present study demonstrates changes in hind limb stance accompanied by changes in the functional properties of muscle afferent neurones in this derangement model of OA. This may provide a possible avenue to explore mechanisms underlying the impaired proprioceptive performance and perhaps other sensory disorders in people with OA.

The role of haptic information in shaping coordination dynamics: inertial frame of reference hypothesis

Current research suggests that non-visual perception of the spatial orientation of body segments is tied to vectors representative of their mass moment distribution (v(mm)). Our question was whether the relative orientation of v(mm) of right and left hands (Δv(mm)=v(mm) left-v(mm) right) constitutes haptic information supporting bimanual coordination and, if so, how it contributes to coordination dynamics. Blindfolded participants coordinated the motions of a pair of cross-shaped, hand-held pendulums that were either symmetrically loaded (Δv(mm)=0) or asymmetrically loaded (Δv(mm)≠0). The sign and magnitude of Δv(mm), in particular of the first moment vector, systematically affected the pattern of coordination (indexed by mean relative phase ϕ), but not its stability. These results suggest that (1) Δv(mm) specifies a frame of reference about which coordination is organized; and (2) that the changes in pattern were a function of the experimentally induced biases in this perceptual frame of reference and not a function of a functional asymmetry akin to detuning. The implications of the findings to the understanding of perceptual regulation of interlimb coordination were discussed.

Principles of part-whole selective perception by dynamic touch extend to the torso

The haptic subsystem of dynamic touch expresses a novel form of part-whole selective perception. When wielding a nonvisible rod grasped at some intermediate point along its length, an individual can attend to and report the length of a part of the rod (e.g., the segment forward of the hand) or the length of the whole rod. Both perceptions relate to the rod's mass moments about the point of grasp but in systematically different ways. Previous demonstrations of this part-whole selectivity have been in respect to rods grasped by hand or attached to a foot. The authors demonstrated the part-whole selectivity for nonvisible rods attached to the shoulder girdle and wielded primarily by movements of the trunk with benchmark performance provided by the same rods grasped and wielded by hand. Their results suggest that part-whole selectivity is a haptic capability general to the body.

Contributions of skin and muscle afferent input to movement sense in the human hand

In the stationary hand, static joint-position sense originates from multimodal somatosensory input (e.g., joint, skin, and muscle). In the moving hand, however, it is uncertain how movement sense arises from these different submodalities of proprioceptors. In contrast to static-position sense, movement sense includes multiple parameters such as motion detection, direction, joint angle, and velocity. Because movement sense is both multimodal and multiparametric, it is not known how different movement parameters are represented by different afferent submodalities. In theory, each submodality could redundantly represent all movement parameters, or, alternatively, different afferent submodalities could be tuned to distinctly different movement parameters. The study described in this paper investigated how skin input and muscle input each contributes to movement sense of the hand, in particular, to the movement parameters dynamic position and velocity. Healthy adult subjects were instructed to indicate with the left hand when they sensed the unseen fingers of the right hand being passively flexed at the metacarpophalangeal (MCP) joint through a previously learned target angle. The experimental approach was to suppress input from skin and/or muscle: skin input by anesthetizing the hand, and muscle input by unexpectedly extending the wrist to prevent MCP flexion from stretching the finger extensor muscle. Input from joint afferents was assumed not to play a significant role because the task was carried out with the MCP joints near their neutral positions. We found that, during passive finger movement near the neutral position in healthy adult humans, both skin and muscle receptors contribute to movement sense but qualitatively differently. Whereas skin input contributes to both dynamic position and velocity sense, muscle input may contribute only to velocity sense.

New method of measuring wrist joint position sense avoiding cutaneous and visual inputs

BACKGROUND

Aspects of afferent inputs, generally termed proprioception, are being increasingly studied. Extraneous factors such as cutaneous inputs can dramatically interfere while trying to design studies in order to determine the participation of the different structures involved in proprioception in the wrist position sense. We tried to determine validity and repeatability of a new wrist joint position measurement device using methodology designed to minimize extraneous factors and isolate muscle and joint inputs.

METHODS

In order to test the reliability of the system, eighty young-adult subjects without musculoskeletal or neurologic impairments affecting the right upper extremity were tested using a custom made motion tracking system. Testing consisted of two conditions: active reproduction of active placement and passive reproduction of passive placement. Subjects performed two repetitions of each target position (10, 20, and 30 degrees of flexion and extension) presented in a random order. Test- retest reliability was then tested.

RESULTS

The average constant error in the passive condition was -0.7 degrees +/- 4.7 degrees as compared to the active condition at 3.7 degrees +/- 5.1 degrees. Average absolute error in the passive condition was 4.9 degrees +/- 2.9 degrees compared to the active condition in which absolute error was 5.9 degrees +/- 3.5 degrees.

DISCUSSION

Test-retest repeatability in both conditions was less than the 5 degrees magnitude typical of clinical goniometry. Errors in the active condition (less than 2 degrees ) were slightly smaller than the passive condition, and the passive condition was also associated with poorer consistency between apparatus sensors and skin sensors.

CONCLUSIONS

The current system for measurement of wrist joint proprioception allows the researcher to decrease extraneous influences that may affect joint position sense awareness, and will help in future study aiming to determine precisely the role of the different structure involved in proprioception.

Where is your arm? Variations in proprioception across space and tasks

The sense of limb position is crucial for movement control and environmental interactions. Our understanding of this fundamental proprioceptive process, however, is limited. For example, little is known about the accuracy of arm proprioception: Does it vary with changes in arm configuration, since some peripheral receptors are engaged only when joints move toward extreme angles? Are these variations consistent across different tasks? Does proprioceptive ability change depending on what we try to localize (e.g., fingertip position vs. elbow angle)? We used a robot exoskeleton to study proprioception in 14 arm configurations across three tasks, asking healthy subjects to 1) match a pointer to elbow angles after passive movements, 2) match a pointer to fingertip positions after passive movements, and 3) actively match their elbow angle to a pointer. Across all three tasks, subjects overestimated more extreme joint positions; this may be due to peripheral sensory signals biasing estimates as a safety mechanism to prevent injury. We also found that elbow angle estimates were more precise when used to judge fingertip position versus directly reported, suggesting that the brain has better access to limb endpoint position than joint angles. Finally, precision of elbow angle estimates improved in active versus passive movements, corroborating work showing that efference copies of motor commands and alpha-gamma motor neuron coactivation contribute to proprioceptive estimates. In sum, we have uncovered fundamental aspects of normal proprioceptive processing, demonstrating not only predictable biases that are dependent on joint configuration and independent of task but also improved precision when integrating information across joints.

A brain-machine interface instructed by direct intracortical microstimulation

Brain–machine interfaces (BMIs) establish direct communication between the brain and artificial actuators. As such, they hold considerable promise for restoring mobility and communication in patients suffering from severe body paralysis. To achieve this end, future BMIs must also provide a means for delivering sensory signals from the actuators back to the brain. Prosthetic sensation is needed so that neuroprostheses can be better perceived and controlled. Here we show that a direct intracortical input can be added to a BMI to instruct rhesus monkeys in choosing the direction of reaching movements generated by the BMI. Somatosensory instructions were provided to two monkeys operating the BMI using either: (a) vibrotactile stimulation of the monkey's hands or (b) multi-channel intracortical microstimulation (ICMS) delivered to the primary somatosensory cortex (S1) in one monkey and posterior parietal cortex (PP) in the other. Stimulus delivery was contingent on the position of the computer cursor: the monkey placed it in the center of the screen to receive machine–brain recursive input. After 2 weeks of training, the same level of proficiency in utilizing somatosensory information was achieved with ICMS of S1 as with the stimulus delivered to the hand skin. ICMS of PP was not effective. These results indicate that direct, bi-directional communication between the brain and neuroprosthetic devices can be achieved through the combination of chronic multi-electrode recording and microstimulation of S1. We propose that in the future, bidirectional BMIs incorporating ICMS may become an effective paradigm for sensorizing neuroprosthetic devices.

Discharges in human muscle receptor afferents during block grasping

Human grasping relies on feedforward control that is monitored and corrected on-line by means of sensory feedback. While much of the sensory mechanisms underpinning hand-object interaction are known, information has been lacking about muscle receptor responses during the phases before and after actual object contact. We therefore let subjects use their thumb and fingers to grasp blocks presented to them while we recorded muscle afferents from the thumb and finger extensor muscles along with wrist and digit kinematics, and electromyographic activity. The kinematics of the task was indistinguishable from "normal" grasping. None of the afferents encoded either object contact or finger apposition. Both primary and secondary afferents were more phase advanced on the parent muscle lengths than expected from previous studies as well as from their responses to imposed length changes of their parent muscles. Thus, the discharges of both primary and secondary afferents were well correlated to the tendon velocity of their parent muscles and that of primary afferents also to acceleration whereas neither appeared to encode muscle length as such. Decoding the velocity of muscle length changes were significantly improved if the discharge of Golgi tendon organ afferents were taken into account along with that of the muscle spindle afferents. We propose that these findings may be explained by the biomechanical properties of contracting muscles. Moreover, we conclude that it seems unlikely that the muscle spindle afferents recorded in this task have any role in providing "proprioceptive" information pertaining to the size of an object grasped.

Fusimotor drive may adjust muscle spindle feedback to task requirements in humans

The aim of the present study was to investigate whether the fusimotor control of muscle spindle sensitivity may depend on the movement parameter the task is focused on, either the velocity or the final position reached. The unitary activities of 18 muscle spindle afferents were recorded by microneurography at the common peroneal nerve. We compared in two situations the responses of muscle spindle afferents to ankle movements imposed while the subject was instructed not to pay attention to or to pay attention to the movement, both in the absence of visual cues. In the two situations, three ramp-and-hold movements were imposed in random order. In one situation, the three movements differed by their velocity and in the other by the final position reached. The task consisted in ranking the three movements according to the parameter under consideration (for example, slow, fast, and medium). The results showed that paying attention to movement velocity gave rise to a significant increase in the dynamic and static responses of muscle afferents. In contrast, focusing attention on the final position reached made the muscle spindle feedback better discriminate the different positions and depressed its capacity to discriminate movement velocities. Changes are interpreted as reflecting dynamic and static gamma activation, respectively. The present results support the view that the fusimotor drive depends on the parameter the task is focused on, so that the muscle afferent feedback is adjusted to the task requirements.

Low-threshold afferent signalling of viscous loads during voluntary movements of the human digits

Humans can discriminate changes in load viscosity during voluntary contractions. The afferent signal origin is unknown. Microneurographic recordings from 83 single low-threshold afferents were made while participants performed triangular ramps either unloaded or with a viscous load. The neural discharges for each cycle were compared across load and velocity. Fifty-eight afferents did not respond. Afferents with sufficient activity were classified as ambiguous--discharges correlated to velocity and load (n=4), infinite viscosity--strong load and weak velocity signal (n=6), no viscosity--strong velocity and weak load signal (n=10) and those with neither (n=5). No single class of afferent provides a coherent signal of viscosity. These data suggest that the central nervous system compares the population response of different inputs to discriminate viscosity.

Afferent input, efference copy, signal noise, and biases in perception of joint angle during active versus passive elbow movements

Psychophysical studies have reported an overestimation of limb position in the direction of movement during the early part of active movements. The main hypothesis tested in this study is that the overestimation results from a process of forward prediction of limb state driven by an efference copy of the outgoing motor command. This hypothesis predicts that position overestimation should decrease or disappear during passive movements, for which there should be no efference copy. Seven subjects were asked to remember and to report the perceived angle of their elbow joint at different times during active and passive movements. They showed a highly velocity-dependent overestimation of the elbow joint angle near the beginning of the movement in both active and passive trials. Toward the end of the movement, subjects showed a relatively velocity-independent underestimation of their elbow angle in all trials. Contrary to the prediction of the efference copy hypothesis, the amplitude and the velocity-dependent slope of the elbow angle overestimation were both greater during the early part of passive movements than active movements. This indicates that psychophysical evidence of early overestimation of arm position on its own is not a sufficient proof of forward prediction based on an efference copy, at least under the conditions of this study. Decreased errors during active movements suggest that an efference copy can improve the accuracy of state estimation during active movements. Error patterns seem to parallel the likely level of sensorimotor noise, suggesting a probabilistic mechanism for position estimation.

Perception of limb orientation in the vertical plane depends on center of mass rather than inertial eigenvectors

We performed two experiments to test the hypothesis that the perception of limb orientation depends on inertial eigenvectors (e_i) against the alternative hypothesis that it depends on the center of mass vector (CM). Whereas e_i constrains the dynamic torques involved in angular rotation, CM constrains the static torque necessary to keep the limb aloft in the gravitational field. Hence, possible effects of e_i and CM on kinesthetic judgments must be related to the dynamic and static torques, respectively, involved in moving and positioning a limb. In the first experiment, blindfolded participants matched, with upper arms supported, the orientation of their forearms while the forearms’ e_i and CM were manipulated relative to the elbow. The manipulation of the vector CM alone induced a matching bias, as did the combined manipulation of e_i and CM, whereas the manipulation of e_i alone did not. In the second experiment, participants positioned their unseen and unsupported right arm at an indicated spatial configuration while e_i and CM of the right forearm were manipulated as in Experiment 1. As in the first experiment, forearm positioning was affected by the independent manipulation of CM and the combined manipulation of e_i and CM, but not by the independent variation of e_i. Moreover, none of the manipulations affected upper arm positioning. These results refute the claim that the perception of limb orientation (in the vertical plane) is based on e_i and demonstrate, for the first time, the implication of a limb segment’s CM in the perception of its orientation.

Cutaneous afferents provide a neuronal population vector that encodes the orientation of human ankle movements

The aim of this study was to analyse the directional coding of two-dimensional limb movements by cutaneous afferents from skin areas covering a multidirectional joint, the ankle. The activity of 89 cutaneous afferents was recorded in the common peroneal nerve, and the mean discharge frequency of each unit was measured during the outward phase of ramp and hold movements imposed in 16 different directions. Forty-two afferents responded to the movements in the following decreasing order (SA2, n = 24/27; FA2, n = 13/17; FA1, n = 3/24; SA1, n = 2/21). All the units activated responded to a specific range of directions, defining their 'preferred sector', within which their response peaked in a given direction, their 'preferred direction'. Based on the distribution of the preferred directions, two populations of afferents, and hence two skin areas were defined: the anterior and the external lateral parts of the leg. As the directional tuning of each population was cosine shaped, the neuronal population vector model was applied and found to efficiently describe the movement direction encoded by cutaneous afferents, as it has been previously reported for muscle afferents. The responses of cutaneous afferents were then considered with respect to those of the afferents from the underlying muscles, which were previously investigated, and an almost perfect matching of directional sensitivity was observed. It is suggested that the common movement-encoding characteristics exhibited by cutaneous and muscle afferents, as early as the peripheral level, may facilitate the central co-processing of their feedbacks subserving kinaesthesia.

Sensorimotor disturbances in chronic neck pain--range of motion, peak velocity, smoothness of movement, and repositioning acuity

The purpose of this pilot study was to evaluate sensorimotor functions in patients with chronic neck pain with objective and quantitative methods. A group of 16 patients with chronic idiopathic neck pain of insidious onset or whiplash associated disorders (WAD) was compared to an equally sized group of healthy subjects. Kinematics were investigated during voluntary head rotations by measuring range of motion, variability of range of motion (ROM-Variability), peak velocity, and smoothness of movement (jerk index). Repositioning acuity after cervical rotations was evaluated by analysing constant and variable error (VE). In comparison to the healthy subjects, the patients showed significantly larger jerk index, ROM-Variability and VE. No statistically significant differences were found between insidious neck pain and WAD. It is concluded that jerky and irregular cervical movements and poor position sense acuity are characteristic sensorimotor symptoms in chronic neck pain. The observed individuality in sensorimotor disturbances emphasizes the importance of developing specific rehabilitation programs for specific dysfunctions, and of using objective and quantitative methods for evaluation of rehabilitation.

Effects of changes in hip position on actions of spinal inhibitory interneurons in humans

The aim of the present study was to establish whether in healthy human subjects the actions of group I muscle afferents arising from the same spinal segments as the soleus innervation (e.g., common peroneal nerve; CPN) or from more proximal spinal segments (femoral nerve; FN) on the soleus H-reflex are modified by changes in hip position. Control and conditioned soleus H-reflexes were elicited and recorded via conventional methods. In seated subjects, CPN and FN stimulation resulted in similar effects to the soleus H-reflex to that previously reported in healthy subjects. However, during hip angle changes, CPN stimulation at the C-T interval of 2 ms resulted in soleus H-reflex depression only when the hip was flexed at 30 degrees , whereas with the hip flexed or extended at 10 degrees the H-reflex was facilitated. CPN stimulation delivered at 100 ms also induced soleus H-reflex facilitation regardless of the hip angle tested. The heteronymous reflex facilitation (conditioned H-reflex with FN stimulation) did not vary systematically with hip angle changes. These findings indicate that hip proprioceptors interact with spinal inhibitory interneurons to enhance spinal reflex excitability under static conditions. This neural switch might constitute an important feature of movement regulation in humans.

Sensory processing during kinesthetic aftereffect following illusory hand movement elicited by tendon vibration

We investigated how the human sensory-motor system elicits a somatosensory aftereffect. Tendon vibration of a limb excites the muscle spindle afferents that contribute to eliciting illusory movements of the limb. After the cessation of vibration, a transient sensation in which the vibrated limb returns towards its original position (kinesthetic aftereffect) is often experienced, even in the absence of the afferent inputs recruited by the vibration. We vibrated the tendon of either the right wrist extensor or flexor muscle that elicited an illusory flexion or extension movement, which was followed by its corresponding extension or flexion aftereffect. First, we psychophysically investigated how the preceding illusory movement affects the aftereffect. Second, we examined the cortico-spinal excitability during the aftereffect to evaluate its changes from the time during the illusion. We measured the amplitude of the motor-evoked potential that is evoked by a single-pulse transcranial magnetic stimulation to the hand section of the contralateral motor cortex (M1). All 19 subjects experienced the aftereffect, and the amount of aftereffect was approximately 70% of the preceding illusion. During the illusion, the cortico-spinal excitability increased more in non-vibrated than in vibrated muscle, so as to reflect the illusory directions. During the aftereffect, the excitability was significantly reduced only in the non-vibrated muscle, with no change in the vibrated muscle, which, in turn, caused an opposite pattern in the unbalanced excitability between the two muscles, and the degree of unbalanced excitability was correlated with the sensation of aftereffect. The kinesthetic aftereffect seems to be elicited by a sensory process that is determined by the preceding illusory movements. Motor-cortical processing of the unbalanced sensory information from the stimulated and non-stimulated muscles may contribute to the elicitation of kinesthetic aftereffect.

The effect of weight-bearing exercise with low frequency, whole body vibration on lumbosacral proprioception: a pilot study on normal subjects

Patients with low back pain (LBP) often present with impaired proprioception of the lumbopelvic region. For this reason, proprioception training usually forms part of the rehabilitation protocols. New exercise equipment that produces whole body, low frequency vibration (WBV) has been developed to improve muscle function, and reportedly improves proprioception. The aim of this pilot study was to investigate whether weightbearing exercise given in conjunction with WBV would affect lumbosacral position sense in healthy individuals. For this purpose, twenty-five young individuals with no LBP were assigned randomly to an experimental or control group. The experimental group received WBV for five minutes while holding a static, semi-squat position. The control group adopted the same weightbearing position for equal time but received no vibration. A two-dimensional motion analysis system measured the repositioning accuracy of pelvic tilting in standing. The experimental (WBV) group demonstrated a significant improvement in repositioning accuracy over time (mean 0.78 degrees) representing 39% improvement. It was concluded that WBV may induce improvements in lumbosacral repositioning accuracy when combined with a weightbearing exercise. Future studies with WBV should focus on evaluating its effects with different types of exercise, the exercise time needed for optimal outcomes, and the effects on proprioception deficits in LBP patients.

The Ia afferent feedback of a given movement evokes the illusion of the same movement when returned to the subject via muscle tendon vibration

The aim of the present study was to further investigate the contribution of primary muscle spindle feedback to proprioception and higher brain functions, such as movement trajectory recognition. For this purpose, complex illusory movements were evoked in subjects by applying patterns of muscle tendon vibration mimicking the natural Ia afferent pattern. Ia afferent messages were previously recorded using microneurographic method from the six main muscle groups acting on the ankle joint during imposed "writing like" movements. The mean Ia afferent pattern was calculated for each muscle group and used as a template to pilot each vibrator. Eleven different vibratory patterns were applied to ten volunteers. Subjects were asked both to copy the perceived illusory movements by hand on a digitizing tablet and to recognize and name the corresponding graphic symbol. The results show that the Ia afferent feedback of a given movement evokes the illusion of the same movement when it is applied to the subject via the appropriate pattern of muscle tendon vibration. The geometry and the kinematic parameters of the imposed and illusory movements are very similar and the so-called "two-thirds power law" is present in the reproduction of the vibration-induced illusory movements. Vibrations within the "natural" frequency range of Ia fibres firing (around 30 Hz) produce clear illusions of movements in all the tested subjects. In addition, increasing the mean frequency of the vibration patterns resulted in a linear increase in the size of the illusory movements. Lastly, the subjects were able to recognize and name the symbols evoked by the vibration-induced primary muscle spindle afferent patterns in 83% of the trials. These findings suggest that the "proprioceptive signature" of a given movement is associated with the corresponding "perceptual signature". The neural mechanisms possibly underlying the sensory to perceptual transformation are discussed in the general framework of "the neuronal population vector model".

Modulation of flexion reflex induced by hip angle changes in human spinal cord injury

The flexion reflex can be elicited via stimulation of skin, muscle, and high-threshold afferents inducing a generalized flexion of the limb. In spinalized animal models this reflex is quite prominent and is strongly modulated by actions of hip proprioceptors. However, analogous actions on the flexion reflex in spinal cord injured (SCI) humans have not yet been examined. In this study, we investigated the effects of imposed static hip angle changes on the flexion reflex in ten motor incomplete SCI subjects when input from plantar cutaneous mechanoreceptors was also present. Flexion reflexes were elicited by low-intensity stimulation of the sural nerve at the lateral malleolus, and were recorded from the ipsilateral tibialis anterior (TA) muscle. Plantar skin stimulation was delivered through two surface electrodes placed on the metatarsals, and was initiated at different delays ranging from 3 to 90 ms. We found that non-noxious sural nerve stimulation induced two types of flexion reflexes in the TA muscle, an early, and a late response. The first was observed only in three subjects and even in these subjects, it appeared irregularly. In contrast, the second (late) flexion reflex was present uniformly in all ten subjects and was significantly modulated during hip angle changes. Flexion reflexes recorded with hip positioned at different angles were compared to the associated control reflexes recorded with hip flexed at 10 degrees. Hip flexion (30 degrees, 40 degrees) depressed the late flexion reflex, while no significant effects were observed with the hip set in neutral angle (0 degrees). Strong facilitatory effects on the late flexion reflex were observed with the hip extended to 10 degrees. Moreover, the effects of plantar skin stimulation on the flexion reflex were also found to depend on the hip angle. The results suggest that hip proprioceptors and plantar cutaneous mechanoreceptors strongly modulate flexion reflex pathways in chronic human SCI, verifying that this type of sensory afferent feedback interact with spinal interneuronal circuits that have been considered as forerunners of stepping and locomotion. The sensory consequences of this afferent input should be considered in rehabilitation programs aimed to restore movement and sensorimotor function in these patients.

Effects of hip joint angle changes on intersegmental spinal coupling in human spinal cord injury

Pathological expression of movement and muscle tone in human upper motor neuron disorders has been partly associated with impaired modulation of spinal inhibitory mechanisms, such as reciprocal or presynaptic inhibition. In addition, input from specific afferent systems contributes significantly to spinal reflex circuits coupled with posture or locomotion. Accordingly, the objectives of this study were to identify the involved afferents and their relative contribution to soleus H-reflex modulation induced by changes in hip position, and to relate these effects with activity of spinal interneuronal circuits. Specifically, we investigated the actions of group I synergistic and antagonistic muscle afferents (e.g. common peroneal nerve, CPN; medial gastrocnemius, MG) and tactile plantar cutaneous afferents on the soleus H-reflex during controlled hip angle variations in 11 motor incomplete spinal cord injured (SCI) subjects. It has been postulated in healthy subjects that CPN stimulation evokes an inhibition on the soleus H-reflex at a conditioning test (C-T) interval of 2-4 ms. This short latency reflex depression is caused mainly by activation of the reciprocal Ia inhibitory pathway. At longer C-T intervals (beyond 30 ms) the soleus H-reflex is again depressed, and is generally accepted to be caused by presynaptic inhibition of soleus Ia afferents. Similarly, MG nerve stimulation depresses soleus H-reflex excitability at the C-T interval of 6 ms, involving the pathway of non-reciprocal group I inhibition, while excitation of plantar cutaneous afferents affects the activity of spinal reflex pathways in the extensors. In this study, soleus H-reflexes recorded alone or during CPN stimulation at either short (2, 3, 4 ms) or long (80, 100, 120 ms) C-T intervals, and MG nerve stimulation delivered at 6 ms were elicited via conventional methods and similar to those adopted in studies conducted in healthy subjects. Plantar skin conditioning stimulation was delivered through two surface electrodes placed on the metatarsals at different C-T intervals ranging from 3 to 90 ms. CPN stimulation at either short or long C-T intervals and MG nerve stimulation resulted in a significant facilitation of the soleus H-reflex, regardless of the hip angle tested. Plantar skin stimulation delivered with hip extended at 10 degrees induced a bimodal facilitation reflex pattern, while with hip flexed (10 degrees , 30 degrees ) the reflex facilitation increased with increments in the C-T interval. This study provides evidence that in human chronic SCI, classically key inhibitory reflex actions are switched to facilitatory, and that spinal processing of plantar cutaneous sensory input and actions of synergistic/antagonistic muscle afferents interact with hip proprioceptive input to facilitate soleus H-reflex excitability. These actions might be associated with the pathological expression of neural control of movement in individuals with SCI, and potentially could be considered in rehabilitation programs geared to restore sensorimotor function in these patients.

Proprioceptive feedback in humans expresses motor invariants during writing

Proprioceptive feedback from populations of muscle spindle afferents feeds the brain with information relating to the instantaneous velocity and direction of ongoing movements. In this paper, we investigate whether the invariant relationship between the velocity and curvature of a trajectory, i.e. the two-thirds power law, is reflected in this muscle spindle feedback. Sixty unitary muscle spindle afferents from six ankle muscle groups were recorded using intraneural microelectrodes during imposed "writing-like" movements. The movements had kinematic parameters obeying the two-thirds power law and were imposed so that the tip of the foot followed trajectories forming four different letters and six numbers. The responses of the muscle spindle afferent populations were analysed using the population vector model. The results demonstrate that the neuronal trajectories attained from populations of muscle spindles clearly depict the path and kinematic parameters and express the movement invariants, i.e. the trajectory segmentation into units of action and the two-thirds power law. The central vs peripheral origin of such constraints involved in the motor system is discussed.

Coding of position by simultaneously recorded sensory neurones in the cat dorsal root ganglion

Muscle, cutaneous and joint afferents continuously signal information about the position and movement of individual joints. How does the nervous system extract more global information, for example about the position of the foot in space? To study this question we used microelectrode arrays to record impulses simultaneously from up to 100 discriminable nerve cells in the L6 and L7 dorsal root ganglia (DRG) of the anaesthetized cat. When the hindlimb was displaced passively with a random trajectory, the firing rate of the neurones could be predicted from a linear sum of positions and velocities in Cartesian (x, y), polar or joint angular coordinates. The process could also be reversed to predict the kinematics of the limb from the firing rates of the neurones with an accuracy of 1-2 cm. Predictions of position and velocity could be combined to give an improved fit to limb position. Decoders trained using random movements successfully predicted cyclic movements and movements in which the limb was displaced from a central point to various positions in the periphery. A small number of highly informative neurones (6-8) could account for over 80% of the variance in position and a similar result was obtained in a realistic limb model. In conclusion, this work illustrates how populations of sensory receptors may encode a sense of limb position and how the firing of even a small number of neurones can be used to decode the position of the limb in space.

"Proprioceptive signature" of cursive writing in humans: a multi-population coding

The goal of the present study was to investigate the firing behavior of populations of muscle spindle afferents in all the muscles acting on the ankle while this joint was being subjected to "writing-like" movements. First it was proposed to determine whether the ensemble of muscle spindles give rise to a unique, specific, and reproducible feedback information characterizing each letter, number or short word. Secondly, we analyzed how the proprioceptive feedback on the whole encodes the spatial and temporal characteristics of writing movements using the "vector population model". The unitary activity of 51 primary and secondary muscle spindle afferents was recorded in the tibial and common peroneal nerves at the level of the popliteal fossea, using the microneurographic method. The units recorded from belonged to the tibialis anterior, the extensor digitorum longus, the extensor hallucis longus, the peroneus lateralis, the gastrocnemius-soleus and the tibialis posterior muscles. The "writing-like" movements were randomly imposed at a "natural" velocity via a computer-controlled machine in a two-dimensional space. In general, muscle spindle afferents from any of the six muscles responded according to the tuning properties of the parent muscle, i.e. increasing their discharge rate during the phases where the direction of movement was within the preferred sensory sector of the parent muscle. The whole trajectory of the writing movements was coded in turn by the activity of Ia afferents arising from all the muscles acting on the joint. Both single afferent responses and population responses were found to be highly specific and reproducible with each graphic sign. The complex multi-muscle afferent pattern involved, with its timing and distribution in the muscle space, seems to constitute a true "proprioceptive signature" for each graphic symbol. The ensemble of muscle spindle afferents were therefore found to encode the instantaneous direction and velocity of writing movements remarkably accurately. It was concluded that the proprioceptive feedback from all the muscles with which the moving joint is equipped provides the CNS with highly specific information that might contribute to a graphic sign identification process.