Limb-state feedback from ensembles of simultaneously recorded dorsal root ganglion neurons

Decoding Bilateral Hindlimb Kinematics From Cat Spinal Signals Using Three-Dimensional Convolutional Neural Network

To date, decoding limb kinematic information mostly relies on neural signals recorded from the peripheral nerve, dorsal root ganglia (DRG), ventral roots, spinal cord gray matter, and the sensorimotor cortex. In the current study, we demonstrated that the neural signals recorded from the lateral and dorsal columns within the spinal cord have the potential to decode hindlimb kinematics during locomotion. Experiments were conducted using intact cats. The cats were trained to walk on a moving belt in a hindlimb-only condition, while their forelimbs were kept on the front body of the treadmill. The bilateral hindlimb joint angles were decoded using local field potential signals recorded using a microelectrode array implanted in the dorsal and lateral columns of both the left and right sides of the cat spinal cord. The results show that contralateral hindlimb kinematics can be decoded as accurately as ipsilateral kinematics. Interestingly, hindlimb kinematics of both legs can be accurately decoded from the lateral columns within one side of the spinal cord during hindlimb-only locomotion. The results indicated that there was no significant difference between the decoding performances obtained using neural signals recorded from the dorsal and lateral columns. The results of the time-frequency analysis show that event-related synchronization (ERS) and event-related desynchronization (ERD) patterns in all frequency bands could reveal the dynamics of the neural signals during movement. The onset and offset of the movement can be clearly identified by the ERD/ERS patterns. The results of the mutual information (MI) analysis showed that the theta frequency band contained significantly more limb kinematics information than the other frequency bands. Moreover, the theta power increased with a higher locomotion speed.

Control of Mammalian Locomotion by Somatosensory Feedback

When animals walk overground, mechanical stimuli activate various receptors located in muscles, joints, and skin. Afferents from these mechanoreceptors project to neuronal networks controlling locomotion in the spinal cord and brain. The dynamic interactions between the control systems at different levels of the neuraxis ensure that locomotion adjusts to its environment and meets task demands. In this article, we describe and discuss the essential contribution of somatosensory feedback to locomotion. We start with a discussion of how biomechanical properties of the body affect somatosensory feedback. We follow with the different types of mechanoreceptors and somatosensory afferents and their activity during locomotion. We then describe central projections to locomotor networks and the modulation of somatosensory feedback during locomotion and its mechanisms. We then discuss experimental approaches and animal models used to investigate the control of locomotion by somatosensory feedback before providing an overview of the different functional roles of somatosensory feedback for locomotion. Lastly, we briefly describe the role of somatosensory feedback in the recovery of locomotion after neurological injury. We highlight the fact that somatosensory feedback is an essential component of a highly integrated system for locomotor control. © 2021 American Physiological Society. Compr Physiol 11:1-71, 2021.

Epidural spinal cord stimulation as an intervention for motor recovery after motor complete spinal cord injury

Spinal cord injury (SCI) commonly results in permanent loss of motor, sensory, and autonomic function. Recent clinical studies have shown that epidural spinal cord stimulation may provide a beneficial adjunct for restoring lower extremity and other neurological functions. Herein, we review the recent clinical advances of lumbosacral epidural stimulation for restoration of sensorimotor function in individuals with motor complete SCI and we discuss the putative neural pathways involved in this promising neurorehabilitative approach. We focus on three main sections: review recent clinical results for locomotor restoration in complete SCI; discuss the contemporary understanding of electrical neuromodulation and signal transduction pathways involved in spinal locomotor networks; and review current challenges of motor system modulation and future directions toward integrative neurorestoration. The current understanding is that initial depolarization occurs at the level of large diameter dorsal root proprioceptive afferents that when integrated with interneuronal and latent residual supraspinal translesional connections can recruit locomotor centers and augment downstream motor units. Spinal epidural stimulation can initiate excitability changes in spinal networks and supraspinal networks. Different stimulation parameters can facilitate standing or stepping, and it may also have potential for augmenting myriad other sensorimotor and autonomic functions. More comprehensive investigation of the mechanisms that mediate the transformation of dysfunctional spinal networks to higher functional states with a greater focus on integrated systems-based control system may reveal the key mechanisms underlying neurological augmentation and motor restoration after severe paralysis.

High-density neural recordings from feline sacral dorsal root ganglia with thin-film array

Objective. Dorsal root ganglia (DRG) are promising sites for recording sensory activity. Current technologies for DRG recording are stiff and typically do not have sufficient site density for high-fidelity neural data techniques.Approach. In acute experiments, we demonstrate single-unit neural recordings in sacral DRG of anesthetized felines using a 4.5µm thick, high-density flexible polyimide microelectrode array with 60 sites and 30-40µm site spacing. We delivered arrays into DRG with ultrananocrystalline diamond shuttles designed for high stiffness affording a smaller footprint. We recorded neural activity during sensory activation, including cutaneous brushing and bladder filling, as well as during electrical stimulation of the pudendal nerve and anal sphincter. We used specialized neural signal analysis software to sort densely packed neural signals.Main results. We successfully delivered arrays in five of six experiments and recorded single-unit sensory activity in four experiments. The median neural signal amplitude was 55μV peak-to-peak and the maximum unique units recorded at one array position was 260, with 157 driven by sensory or electrical stimulation. In one experiment, we used the neural analysis software to track eight sorted single units as the array was retracted ∼500μm.Significance. This study is the first demonstration of ultrathin, flexible, high-density electronics delivered into DRG, with capabilities for recording and tracking sensory information that are a significant improvement over conventional DRG interfaces.

Decoding hindlimb kinematics from descending and ascending neural signals during cat locomotion

OBJECTIVE

The main objective of this research is to record both sensory and motor information from the ascending and descending tracts within the spinal cord for decoding the hindlimb kinematics during walking on the treadmill.

APPROACH

Two different experimental paradigms (i.e., active and passive) were used in the current study. During active experiments, five cats were trained to walk bipedally while their hands kept on the front frame of the treadmill for balance or to walk quadrupedally. During passive experiments, the limb was passively moved by the experimenter. Local field potential (LFP) activity was recorded using a microwire array implanted in the dorsal column (DC) and lateral column (LC) of the L3-L4 spinal segments. The amplitude and frequency components of the LFP formed the feature set and the elastic net regularization was used to decode the hindlimb joint angles.

MAIN RESULTS

The results show that there is no significant difference between the information content of the signals recorded from the DC and LC regions during walking on the treadmill, but the information content of the DC is significantly higher than that of the LC during passively applied movement of the hindlimb in the anesthetized cats. Moreover, the decoding performance obtained using the recorded signals from the DC is comparable with that from the LC during locomotion. But, the decoding performance obtained using the recording channels in the DC is significantly better than that obtained using the signals recorded from the LC. The long-term analysis shows that robust decoding performance can be achieved over 2-3 months without a significant decrease in performance.

SIGNIFICANCE

This work presents a promising approach to developing a natural and robust motor neuroprosthesis device using descending neural signals to execute the movement and ascending neural signals as the feedback information for control of the movement.

Bladder pressure encoding by sacral dorsal root ganglion fibres: implications for decoding

OBJECTIVE

We aim at characterising the encoding of bladder pressure (intravesical pressure) by a population of sensory fibres. This research is motivated by the possibility to restore bladder function in elderly patients or after spinal cord injury using implanted devices, so called bioelectronic medicines. For these devices, nerve-based estimation of intravesical pressure can enable a personalized and on-demand stimulation paradigm, which has promise of being more effective and efficient. In this context, a better understanding of the encoding strategies employed by the body might in the future be exploited by informed decoding algorithms that enable a precise and robust bladder-pressure estimation.

APPROACH

To this end, we apply information theory to microelectrode-array recordings from the cat sacral dorsal root ganglion while filling the bladder, conduct surrogate data studies to augment the data we have, and finally decode pressure in a simple informed approach.

MAIN RESULTS

We find an encoding scheme by different main bladder neuron types that we divide into three response types (slow tonic, phasic, and derivative fibres). We show that an encoding by different bladder neuron types, each represented by multiple cells, offers reliability through within-type redundancy and high information rates through semi-independence of different types. Our subsequent decoding study shows a more robust decoding from mean responses of homogeneous cell pools.

SIGNIFICANCE

We have here, for the first time, established a link between an information theoretic analysis of the encoding of intravesical pressure by a population of sensory neurons to an informed decoding paradigm. We show that even a simple adapted decoder can exploit the redundancy in the population to be more robust against cell loss. This work thus paves the way towards principled encoding studies in the periphery and towards a new generation of informed peripheral nerve decoders for bioelectronic medicines.

Spatial models of cell distribution in human lumbar dorsal root ganglia

Dorsal root ganglia (DRG), which contain the somata of primary sensory neurons, have increasingly been considered as novel targets for clinical neural interfaces, both for neuroprosthetic and pain applications. Effective use of either neural recording or stimulation technologies requires an appropriate spatial position relative to the target neural element, whether axon or cell body. However, the internal three-dimensional spatial organization of human DRG neural fibers and somata has not been quantitatively described. In this study, we analyzed 202 cross-sectional images across the length of 31 human L4 and L5 DRG from 10 donors. We used a custom semi-automated graphical user interface to identify the locations of neural elements in the images and normalize the output to a consistent spatial reference for direct comparison by spinal level. By applying a recursive partitioning algorithm, we found that the highest density of cell bodies at both spinal levels could be found in the inner 85% of DRG length, the outer-most 25-30% radially, and the dorsal-most 69-76%. While axonal density was fairly homogeneous across the DRG length, there was a distinct low density region in the outer 7-11% radially. These findings are consistent with previous qualitative reports of neural distribution in DRG. The quantitative measurements we provide will enable improved targeting of future neural interface technologies and DRG-focused pharmaceutical therapies, and provide a rigorous anatomical description of the bridge between the central and peripheral nervous systems.

DRG microstimulation evokes postural responses in awake, standing felines

Objective.

We have demonstrated previously that microstimulation in the dorsal root ganglia (DRG) can selectively evoke activity in primary afferent neurons in anesthetized cats. This study describes the results of experiments focused on characterizing the postural effects of DRG microstimulation in awake cats during quiet standing.

Approach.

To understand the parameters of stimulation that can affect these postural shifts, we measured changes in ground reaction forces (GRF) while varying stimulation location and amplitude. Four animals were chronically implanted at the L6 and L7 DRG with penetrating multichannel microelectrode arrays. During each week of testing, we identified electrode channels that recruited primary afferent neurons with fast (80–120 m s⁻¹) and medium (30–75 m s⁻¹) conduction velocities, and selected one channel to deliver current-controlled biphasic stimulation trains during quiet standing.

Main results.

Postural responses were identified by changes in GRFs and were characterized based on their magnitude and latency. During DRG microstimulation, animals did not exhibit obvious signs of distress or discomfort, which could be indicative of pain or aversion to a noxious sensation. Across 56 total weeks, 13 electrode channels evoked behavioral responses, as detected by a significant change in GRF. Stimulation amplitude modulated the magnitude of the GRF responses for these 13 channels (p < 0.001). It was not possible to predict whether or not an electrode would drive a behavioral response based on information including conduction velocity, recruitment threshold, or the DRG in which it resided.

Significance.

The distinct and repeatable effects on the postural response to low amplitude (<40 μA) DRG microstimulation support that this technique may be an effective way to restore somatosensory feedback after neurological injuries such as amputation.

Estimation of Bladder Pressure and Volume from the Neural Activity of Lumbosacral Dorsal Horn Using a Long-Short-Term-Memory-based Deep Neural Network

In this paper, we propose a deep recurrent neural network (DRNN) for the estimation of bladder pressure and volume from neural activity recorded directly from spinal cord gray matter neurons. The model was based on the Long Short-Term Memory (LSTM) architecture, which has emerged as a general and effective model for capturing long-term temporal dependencies with good generalization performance. In this way, training the network with the data recorded from one rat could lead to estimating the bladder status of different rats. We combined modeling of spiking and local field potential (LFP) activity into a unified framework to estimate the pressure and volume of the bladder. Moreover, we investigated the effect of two-electrode recording on decoding performance. The results show that the two-electrode recordings significantly improve the decoding performance compared to single-electrode recordings. The proposed framework could estimate bladder pressure and volume with an average normalized root-mean-squared (NRMS) error of 14.9 ± 4.8% and 19.7 ± 4.7% and a correlation coefficient (CC) of 83.2 ± 3.2% and 74.2 ± 6.2%, respectively. This work represents a promising approach to the real-time estimation of bladder pressure/volume in the closed-loop control of bladder function using functional electrical stimulation.

Implantable Neural Interfaces and Wearable Tactile Systems for Bidirectional Neuroprosthetics Systems

Neuroprosthetics and neuromodulation represent a promising field for several related applications in the central and peripheral nervous system, such as the treatment of neurological disorders, the control of external robotic devices, and the restoration of lost tactile functions. These actions are allowed by the neural interface, a miniaturized implantable device that most commonly exploits electrical energy to fulfill these operations. A neural interface must be biocompatible, stable over time, low invasive, and highly selective; the challenge is to develop a safe, compact, and reliable tool for clinical applications. In case of anatomical impairments, neuroprosthetics is bound to the need of exploring the surrounding environment by fast-responsive and highly sensitive artificial tactile sensors that mimic the natural sense of touch. Tactile sensors and neural interfaces are closely interconnected since the readouts from the first are required to convey information to the neural implantable apparatus. The role of these devices is pivotal hence technical improvements are essential to ensure a secure system to be eventually adopted in daily life. This review highlights the fundamental criteria for the design and microfabrication of neural interfaces and artificial tactile sensors, their use in clinical applications, and future enhancements for the release of a second generation of devices.

A review for the peripheral nerve interface designer

Informational density and relative accessibility of the peripheral nervous system make it an attractive site for therapeutic intervention. Electrode-based electrophysiological interfaces with peripheral nerves have been under development since the 1960s and, for several applications, have seen widespread clinical implementation. However, many applications require a combination of neural target resolution and stability which has thus far eluded existing peripheral nerve interfaces (PNIs). With the goal of aiding PNI designers in development of devices that meet the demands of next-generation applications, this review seeks to collect and present practical considerations and best practices which emerge from the literature, including both lessons learned during early PNI development and recent ideas. Fundamental and practical principles guiding PNI design are reviewed, followed by an updated and critical account of existing PNI designs and strategies. Finally, a brief survey of in vitro and in vivo PNI characterization methods is presented.

A probabilistic recurrent neural network for decoding hind limb kinematics from multi-segment recordings of the dorsal horn neurons

OBJECTIVE

Providing accurate and robust estimates of limb kinematics from recorded neural activities is prominent in closed-loop control of functional electrical stimulation (FES). A major issue in providing accurate decoding the limb kinematics is the decoding model. The primary goal of this study is to develop a decoding approach to model the dynamic interactions of neural systems for accurate decoding. Another critical issue is to find reliable recording sites. Up to now, neural recordings from spinal neural activities were investigated. In this paper, the neural recordings from different vertebrae in decoding limb kinematics are investigated.

APPROACH

In the current study, a new generative probabilistic model with explicit considering the joint density is developed. Then, an adaptive discriminative learning algorithm is proposed for learning the model. It will be shown that the proposed generative process can be implemented by a recurrent neural network (RNN) with specific structure. We record the neural activities from dorsal horn neurons by using three electrodes placed in the L4, L5, and L6 vertebrae in anesthetized cats.

MAIN RESULTS

Information theoretic analysis on single-joint movement and multi-segment recordings implies the rostrocaudal distribution of kinematic information. It is demonstrated that during hip movement, best decoding performance is achieved by L4 recordings. For knee and ankle movements, best decodings are achieved by L5, and L6 recordings respectively. It is also shown that the decoding accuracy using multi-segment recordings outperforms decoding accuracy obtained by single-segment recording in multi-joint movement. The results also confirm the superiority of the proposed probabilistic recurrent neural network (PRNN) over the conventional recurrent neural network and Kalman filter ([Formula: see text]).

SIGNIFICANCE

Multi-segment recordings from dorsal horn neurons as well as the proposed probabilistic recurrent network model provide a promising approach for robust and accurate decoding limb kinematics.

On the development of optical peripheral nerve interfaces

Limb loss and spinal cord injury are two debilitating conditions that continue to grow in prevalence. Prosthetic limbs and limb reanimation present two ways of providing affected individuals with means to interact in the world. These techniques are both dependent on a robust interface with the peripheral nerve. Current methods for interfacing with the peripheral nerve tend to suffer from low specificity, high latency and insufficient robustness for a chronic implant. An optical peripheral nerve interface may solve some of these problems by decreasing invasiveness and providing single axon specificity. In order to implement such an interface three elements are required: (1) a transducer capable of translating light into a neural stimulus or translating neural activity into changes in fluorescence, (2) a means for delivering said transducer and (3) a microscope for providing the stimulus light and detecting the fluorescence change. There are continued improvements in both genetically encoded calcium and voltage indicators as well as new optogenetic actuators for stimulation. Similarly, improvements in specificity of viral vectors continue to improve expression in the axons of the peripheral nerve. Our work has recently shown that it is possible to virally transduce axons of the peripheral nerve for recording from small fibers. The improvements of these components make an optical peripheral nerve interface a rapidly approaching alternative to current methods.

Recording single- and multi-unit neuronal action potentials from the surface of the dorsal root ganglion

The dorsal root ganglia (DRG) contain cell bodies of primary afferent neurons, which are frequently studied by recording extracellularly with penetrating microelectrodes inserted into the DRG. We aimed to isolate single- and multi-unit activity from primary afferents in the lumbar DRG using non-penetrating electrode arrays and to characterize the relationship of that activity with limb position and movement. The left sixth and seventh lumbar DRG (L6-L7) were instrumented with penetrating and non-penetrating electrode arrays to record neural activity during passive hindlimb movement in 7 anesthetized cats. We found that the non-penetrating arrays could record both multi-unit and well-isolated single-unit activity from the surface of the DRG, although with smaller signal to noise ratios (SNRs) compared to penetrating electrodes. Across all recorded units, the median SNR was 1.1 for non-penetrating electrodes and 1.6 for penetrating electrodes. Although the non-penetrating arrays were not anchored to the DRG or surrounding tissues, the spike amplitudes did not change (<1% change from baseline spike amplitude) when the limb was moved passively over a limited range of motion (~20 degrees at the hip). Units of various sensory fiber types were recorded, with 20% of units identified as primary muscle spindles, 37% as secondary muscle spindles, and 24% as cutaneous afferents. Our study suggests that non-penetrating electrode arrays can record modulated single- and multi-unit neural activity of various sensory fiber types from the DRG surface.

Proprioceptive Recognition with Artificial Neural Networks Based on Organizations of Spinocerebellar Tract and Cerebellum

Muscle kinematics and kinetics are nonlinearly encoded by proprioceptors, and the changes in muscle length and velocity are integrated into Ia afferent. Besides, proprioceptive signals from multiple muscles are probably mixed in afferent pathways, which all lead to difficulties in proprioceptive recognition for the cerebellum. In this study, the artificial neural networks, whose organizations are biologically based on the spinocerebellar tract and cerebellum, are utilized to decode the proprioceptive signals. Consistent with the controversy of the proprioceptive division in the dorsal spinocerebellar tract, the spinocerebellar tract networks incorporated two distinct inferences, (1) the centralized networks, which mixed Ia, II, and Ib and processed them together; (2) the decentralized networks, which processed Ia, II, and Ib afferents separately. The cerebellar networks were based on the Marr-Albus model to recognize the kinematic states. The networks were trained by a specific movement, and the trained networks were subsequently required to predict kinematic states of six movements. The results demonstrated that the centralized networks, which were more consistent with the physiological findings in recent years, had better recognition accuracy than the decentralized networks, and the networks were still effective even when proprioceptive afferents from multiple muscles were integrated.

A speed-adaptive intraspinal microstimulation controller to restore weight-bearing stepping in a spinal cord hemisection model

OBJECTIVE

The goal of this study was to develop control strategies to produce alternating, weight-bearing stepping in a cat model of hemisection spinal cord injury (SCI) using intraspinal microstimulation (ISMS).

APPROACH

Six cats were anesthetized and the functional consequences of a hemisection SCI were simulated by manually moving one hind-limb through the gait cycle over a moving treadmill belt. ISMS activated the muscles in the other leg by stimulating motor networks in the lumbosacral enlargement using low levels of current (<110 µA). The control strategy used signals from ground reaction forces and angular velocity from the manually-moved limb to anticipate states of the gait cycle, and controlled ISMS to move the other hind-limb into the opposite state. Adaptive control strategies were developed to ensure weight-bearing at different stepping speeds. The step period was predicted using generalizations obtained through four supervised machine learning algorithms and used to adapt the control strategy for faster steps.

MAIN RESULTS

At a single speed, 100% of the steps had sufficient weight-bearing; at faster speeds without adaptation, 97.6% of steps were weight-bearing (significantly less than that for single speed; p  =  0.002). By adapting the control strategy for faster steps using the predicted step period, weight-bearing was achieved in more than 99% of the steps in three of four methods (significantly more than without adaptation p  <  0.04). Overall, a multivariate model tree increased the number of weight-bearing steps, restored step symmetry, and maintained alternation at faster stepping speeds.

SIGNIFICANCE

Through the adaptive control strategies guided by supervised machine learning, we were able to restore weight-bearing and maintain alternation and step symmetry at varying stepping speeds.

Linear feature projection-based real-time decoding of limb state from dorsal root ganglion recordings

Proprioceptive afferent activities recorded by a multichannel microelectrode have been used to decode limb movements to provide sensory feedback signals for closed-loop control in a functional electrical stimulation (FES) system. However, analyzing the high dimensionality of neural activity is one of the major challenges in real-time applications. This paper proposes a linear feature projection method for the real-time decoding of ankle and knee joint angles. Single-unit activity was extracted as a feature vector from proprioceptive afferent signals that were recorded from the L7 dorsal root ganglion during passive movements of ankle and knee joints. The dimensionality of this feature vector was then reduced using a linear feature projection composed of projection pursuit and negentropy maximization (PP/NEM). Finally, a time-delayed Kalman filter was used to estimate the ankle and knee joint angles. The PP/NEM approach had a better decoding performance than did other feature projection methods, and all processes were completed within the real-time constraints. These results suggested that the proposed method could be a useful decoding method to provide real-time feedback signals in closed-loop FES systems.

Linear Feature Projection-Based Sensory Event Detection from the Multiunit Activity of Dorsal Root Ganglion Recordings

Afferent signals recorded from the dorsal root ganglion can be used to extract sensory information to provide feedback signals in a functional electrical stimulation (FES) system. The goal of this study was to propose an efficient feature projection method for detecting sensory events from multiunit activity-based feature vectors of tactile afferent activity. Tactile afferent signals were recorded from the L4 dorsal root ganglion using a multichannel microelectrode for three types of sensory events generated by mechanical stimulation on the rat hind paw. The multiunit spikes (MUSs) were extracted as multiunit activity-based feature vectors and projected using a linear feature projection method which consisted of projection pursuit and negentropy maximization (PP/NEM). Finally, a multilayer perceptron classifier was used to detect sensory events. The proposed method showed a detection accuracy superior to those of other linear and nonlinear feature projection methods and all processes were completed within real-time constraints. Results suggest that the proposed method could be useful to detect sensory events in real time. We have demonstrated the methodology for an efficient feature projection method to detect real-time sensory events from the multiunit activity of dorsal root ganglion recordings. The proposed method could be applied to provide real-time sensory feedback signals in closed-loop FES systems.

Evaluation of Decoding Algorithms for Estimating Bladder Pressure from Dorsal Root Ganglia Neural Recordings

A closed-loop device for bladder control may offer greater clinical benefit compared to current open-loop stimulation devices. Previous studies have demonstrated the feasibility of using single-unit recordings from sacral-level dorsal root ganglia (DRG) for decoding bladder pressure. Automatic online sorting, to differentiate single units, can be computationally heavy and unreliable, in contrast to simple multi-unit thresholded activity. In this study, the feasibility of using DRG multi-unit recordings to decode bladder pressure was examined. A broad range of feature selection methods and three algorithms (multivariate linear regression, basic Kalman filter, and a nonlinear autoregressive moving average model) were used to create training models and provide validation fits to bladder pressure for data collected in seven anesthetized feline experiments. A non-linear autoregressive moving average (NARMA) model with regularization provided the most accurate bladder pressure estimate, based on normalized root-mean-squared error, NRMSE, (17 ± 7%). A basic Kalman filter yielded the highest similarity to the bladder pressure with an average correlation coefficient, CC, of 0.81 ± 0.13. The best algorithm set (based on NRMSE) was further evaluated on data obtained from a chronic feline experiment. Testing results yielded a NRMSE and CC of 10.7% and 0.61, respectively from a model that was trained on data recorded 2 weeks prior. From offline analysis, implementation of NARMA in a closed-loop scheme for detecting bladder contractions would provide a robust control signal. Ultimate integration of closed-loop algorithms in bladder neuroprostheses will require evaluations of parameter and signal stability over time.

Decoding hind limb kinematics from neuronal activity of the dorsal horn neurons using multiple level learning algorithm

Decoding continuous hind limb joint angles from sensory recordings of neural system provides a feedback for closed-loop control of hind limb movement using functional electrical stimulation. So far, many attempts have been done to extract sensory information from dorsal root ganglia and sensory nerves. In this work, we examine decoding joint angles trajectories from the single-electrode extracellular recording of dorsal horn gray matter of the spinal cord during passive limb movement in anesthetized cats. In this study, a processing framework based on ensemble learning approach is propose to combine firing rate (FR) and interspike interval (ISI) information of the neuronal activity. For this purpose, a stacked generalization approach based on recurrent neural network is proposed to enhance decoding accuracy of the movement kinematics. The results show that the high precision neural decoding of limb movement can be achieved even with a single electrode implanted in the spinal cord gray matter.

Neurophysiology and neural engineering: a review

Neurophysiology is the branch of physiology concerned with understanding the function of neural systems. Neural engineering (also known as neuroengineering) is a discipline within biomedical engineering that uses engineering techniques to understand, repair, replace, enhance, or otherwise exploit the properties and functions of neural systems. In most cases neural engineering involves the development of an interface between electronic devices and living neural tissue. This review describes the origins of neural engineering, the explosive development of methods and devices commencing in the late 1950s, and the present-day devices that have resulted. The barriers to interfacing electronic devices with living neural tissues are many and varied, and consequently there have been numerous stops and starts along the way. Representative examples are discussed. None of this could have happened without a basic understanding of the relevant neurophysiology. I also consider examples of how neural engineering is repaying the debt to basic neurophysiology with new knowledge and insight.

Musculoskeletal geometry accounts for apparent extrinsic representation of paw position in dorsal spinocerebellar tract

Proprioception, the sense of limb position and motion, arises from individual muscle receptors. An important question is how and where in the neuroaxis our high level "extrinsic" sense of limb movement originates. In the 1990s, a series of papers detailed the properties of neurons in the dorsal spinocerebellar tract (DSCT) of the cat. Despite their direct projections from sensory receptors, it appeared that half of these neurons had consistent, high-level tuning to paw position rather than to joint angles (or muscle lengths). These results suggested that many DSCT neurons compute paw position from lower level sensory information. We examined the contribution of musculoskeletal geometry to this apparent extrinsic representation by simulating a three-joint hindlimb with mono- and biarticular muscles, each providing a muscle spindlelike signal, modulated by the muscle length. We simulated neurons driven by randomly weighted combinations of these signals and moved the paw to different positions under two joint-covariance conditions similar to the original experiments. Our results paralleled those experiments in a number of respects: 1) Many neurons were tuned to paw position relative to the hip under both conditions. 2) The distribution of tuning was strongly bimodal, with most neurons driven by whole-leg flexion or extension. 3) The change in tuning between conditions clustered around zero (median absolute change ~20°). These results indicate that, at least for these constraint conditions, extrinsic-like representation can be achieved simply through musculoskeletal geometry and convergent muscle length inputs. Consequently, they suggest a reinterpretation of the earlier results may be required.NEW & NOTEWORTHY A classic experiment concluding that many dorsal spinocerebellar tract neurons encode paw position rather than joint angles has been cited by many studies as evidence for high-level computation occurring within a single synapse of the sensors. However, our study provides evidence that such a computation is not required to explain the results. Using simulation, we replicated many of the original results with purely random connectivity, suggesting that a reinterpretation of the classic experiment is needed.

Multiunit Activity-Based Real-Time Limb-State Estimation from Dorsal Root Ganglion Recordings

Proprioceptive afferent activities could be useful for providing sensory feedback signals for closed-loop control during functional electrical stimulation (FES). However, most previous studies have used the single-unit activity of individual neurons to extract sensory information from proprioceptive afferents. This study proposes a new decoding method to estimate ankle and knee joint angles using multiunit activity data. Proprioceptive afferent signals were recorded from a dorsal root ganglion with a single-shank microelectrode during passive movements of the ankle and knee joints, and joint angles were measured as kinematic data. The mean absolute value (MAV) was extracted from the multiunit activity data, and a dynamically driven recurrent neural network (DDRNN) was used to estimate ankle and knee joint angles. The multiunit activity-based MAV feature was sufficiently informative to estimate limb states, and the DDRNN showed a better decoding performance than conventional linear estimators. In addition, processing time delay satisfied real-time constraints. These results demonstrated that the proposed method could be applicable for providing real-time sensory feedback signals in closed-loop FES systems.

Hysteretic behavior of bladder afferent neurons in response to changes in bladder pressure

Background

Mechanosensitive afferents innervating the bladder increase their firing rate as the bladder fills and pressure rises. However, the relationship between afferent firing rates and intravesical pressure is not a simple linear one. Firing rate responses to pressure can differ depending on prior activity, demonstrating hysteresis in the system. Though this hysteresis has been commented on in published literature, it has not been quantified.

Results

Sixty-six bladder afferents recorded from sacral dorsal root ganglia in five alpha-chloralose anesthetized felines were identified based on their characteristic responses to pressure (correlation coefficient ≥ 0.2) during saline infusion (2 ml/min). For saline infusion trials, we calculated a maximum hysteresis ratio between the firing rate difference at each pressure and the overall firing rate range (or Hmax) of 0.86 ± 0.09 (mean ± standard deviation) and mean hysteresis ratio (or Hmean) of 0.52 ± 0.13 (n = 46 afferents). For isovolumetric trials in two experiments (n = 33 afferents) Hmax was 0.72 ± 0.14 and Hmean was 0.40 ± 0.14.

Conclusions

A comprehensive state model that integrates these hysteresis parameters to determine the bladder state may improve upon existing neuroprostheses for bladder control.

Quantitative Analysis of Rat Dorsal Root Ganglion Neurons Cultured on Microelectrode Arrays Based on Fluorescence Microscopy Image Processing

Microelectrode Arrays (MEA) are devices for long term electrophysiological recording of extracellular spontaneous or evocated activities on in vitro neuron culture. This work proposes and develops a framework for quantitative and morphological analysis of neuron cultures on MEAs, by processing their corresponding images, acquired by fluorescence microscopy. The neurons are segmented from the fluorescence channel images using a combination of segmentation by thresholding, watershed transform, and object classification. The positioning of microelectrodes is obtained from the transmitted light channel images using the circular Hough transform. The proposed method was applied to images of dissociated culture of rat dorsal root ganglion (DRG) neuronal cells. The morphological and topological quantitative analysis carried out produced information regarding the state of culture, such as population count, neuron-to-neuron and neuron-to-microelectrode distances, soma morphologies, neuron sizes, neuron and microelectrode spatial distributions. Most of the analysis of microscopy images taken from neuronal cultures on MEA only consider simple qualitative analysis. Also, the proposed framework aims to standardize the image processing and to compute quantitative useful measures for integrated image-signal studies and further computational simulations. As results show, the implemented microelectrode identification method is robust and so are the implemented neuron segmentation and classification one (with a correct segmentation rate up to 84%). The quantitative information retrieved by the method is highly relevant to assist the integrated signal-image study of recorded electrophysiological signals as well as the physical aspects of the neuron culture on MEA. Although the experiments deal with DRG cell images, cortical and hippocampal cell images could also be processed with small adjustments in the image processing parameter estimation.

Cortical neuroprosthetics from a clinical perspective

Recent pilot clinical studies have demonstrated that subjects with severe disorders of movement and communication can exert direct neural control over assistive devices using invasive Brain-Machine Interface (BMI) technology, also referred to as 'cortical neuroprosthetics'. These important proof-of-principle studies have generated great interest among those with disability and clinicians who provide general medical, neurological and/or rehabilitative care. Taking into account the perspective of providers who may be unfamiliar with the field, we first review the clinical goals and fundamentals of invasive BMI technology, and then briefly summarize the vast body of basic science research demonstrating its feasibility. We emphasize recent translational progress in the target clinical populations and discuss translational challenges and future directions.

Sensory control of normal movement and of movement aided by neural prostheses

Signals from sensory receptors in muscles and skin enter the central nervous system (CNS), where they contribute to kinaesthesia and the generation of motor commands. Many lines of evidence indicate that sensory input from skin receptors, muscle spindles and Golgi tendon organs play the predominant role in this regard. Yet in spite of over 100 years of research on this topic, some quite fundamental questions remain unresolved. How does the CNS choose to use the ability to control muscle spindle sensitivity during voluntary movements? Do spinal reflexes contribute usefully to load compensation, given that the feedback gain must be quite low to avoid instability? To what extent do signals from skin stretch receptors contribute? This article provides a brief review of various theories, past and present, that address these questions. To what extent has the knowledge gained resulted in clinical applications? Muscles paralyzed as a result of spinal cord injury or stroke can be activated by electrical stimulation delivered by neuroprostheses. In practice, at most two or three sensors can be deployed on the human body, providing only a small fraction of the information supplied by the tens of thousands of sensory receptors in animals. Most of the neuroprostheses developed so far do not provide continuous feedback control. Instead, they switch from one state to another when signals from their one or two sensors meet pre-set thresholds (finite state control). The inherent springiness of electrically activated muscle provides a crucial form of feedback control that helps smooth the resulting movements. In spite of the dissimilarities, parallels can be found between feedback control in neuroprostheses and in animals and this can provide surprising insights in both directions.

Evaluation of poly(3,4-ethylenedioxythiophene)/carbon nanotube neural electrode coatings for stimulation in the dorsal root ganglion

OBJECTIVE

The dorsal root ganglion is an attractive target for implanting neural electrode arrays that restore sensory function or provide therapy via stimulation. However, penetrating microelectrodes designed for these applications are small and deliver low currents. For long-term performance of microstimulation devices, novel coating materials are needed in part to decrease impedance values at the electrode-tissue interface and to increase charge storage capacity.

APPROACH

Conductive polymer poly(3,4-ethylenedioxythiophene) (PEDOT) and multi-wall carbon nanotubes (CNTs) were coated on the electrode surface and doped with the anti-inflammatory drug, dexamethasone. Electrode characteristics and the tissue reaction around neural electrodes as a result of stimulation, coating and drug release were characterized. Hematoxylin and eosin staining along with antibodies recognizing Iba1 (microglia/macrophages), NF200 (neuronal axons), NeuN (neurons), vimentin (fibroblasts), caspase-3 (cell death) and L1 (neural cell adhesion molecule) were used. Quantitative image analyses were performed using MATLAB.

MAIN RESULTS

Our results indicate that coated microelectrodes have lower in vitro and in vivo impedance values. Significantly less neuronal death/damage was observed with coated electrodes as compared to non-coated controls. The inflammatory response with the PEDOT/CNT-coated electrodes was also reduced.

SIGNIFICANCE

This study is the first to report on the utility of these coatings in stimulation applications. Our results indicate PEDOT/CNT coatings may be valuable additions to implantable electrodes used as therapeutic modalities.

Robotics, stem cells, and brain-computer interfaces in rehabilitation and recovery from stroke: updates and advances

OBJECTIVE

The aim of this study was to describe the current state and latest advances in robotics, stem cells, and brain-computer interfaces in rehabilitation and recovery for stroke.

DESIGN

The authors of this summary recently reviewed this work as part of a national presentation. The article represents the information included in each area.

RESULTS

Each area has seen great advances and challenges as products move to market and experiments are ongoing.

CONCLUSIONS

Robotics, stem cells, and brain-computer interfaces all have tremendous potential to reduce disability and lead to better outcomes for patients with stroke. Continued research and investment will be needed as the field moves forward. With this investment, the potential for recovery of function is likely substantial.

Sensory systems in the control of movement

Animal movement is immensely varied, from the simplest reflexive responses to the most complex, dexterous voluntary tasks. Here, we focus on the control of movement in mammals, including humans. First, the sensory inputs most closely implicated in controlling movement are reviewed, with a focus on somatosensory receptors. The response properties of the large muscle receptors are examined in detail. The role of sensory input in the control of movement is then discussed, with an emphasis on the control of locomotion. The interaction between central pattern generators and sensory input, in particular in relation to stretch reflexes, timing, and pattern forming neuronal networks is examined. It is proposed that neural signals related to bodily velocity form the basic descending command that controls locomotion through specific and well-characterized relationships between muscle activation, step cycle phase durations, and biomechanical outcomes. Sensory input is crucial in modulating both the timing and pattern forming parts of this mechanism.

Real-time control of walking using recordings from dorsal root ganglia

OBJECTIVE

The goal of this study was to decode sensory information from the dorsal root ganglia (DRG) in real time, and to use this information to adapt the control of unilateral stepping with a state-based control algorithm consisting of both feed-forward and feedback components.

APPROACH

In five anesthetized cats, hind limb stepping on a walkway or treadmill was produced by patterned electrical stimulation of the spinal cord through implanted microwire arrays, while neuronal activity was recorded from the DRG. Different parameters, including distance and tilt of the vector between hip and limb endpoint, integrated gyroscope and ground reaction force were modelled from recorded neural firing rates. These models were then used for closed-loop feedback.

MAIN RESULTS

Overall, firing-rate-based predictions of kinematic sensors (limb endpoint, integrated gyroscope) were the most accurate with variance accounted for >60% on average. Force prediction had the lowest prediction accuracy (48 ± 13%) but produced the greatest percentage of successful rule activations (96.3%) for stepping under closed-loop feedback control. The prediction of all sensor modalities degraded over time, with the exception of tilt.

SIGNIFICANCE

Sensory feedback from moving limbs would be a desirable component of any neuroprosthetic device designed to restore walking in people after a spinal cord injury. This study provides a proof-of-principle that real-time feedback from the DRG is possible and could form part of a fully implantable neuroprosthetic device with further development.

Neuroprosthetic technology for individuals with spinal cord injury

CONTEXT

Spinal cord injury (SCI) results in a loss of function and sensation below the level of the lesion. Neuroprosthetic technology has been developed to help restore motor and autonomic functions as well as to provide sensory feedback.

FINDINGS

This paper provides an overview of neuroprosthetic technology that aims to address the priorities for functional restoration as defined by individuals with SCI. We describe neuroprostheses that are in various stages of preclinical development, clinical testing, and commercialization including functional electrical stimulators, epidural and intraspinal microstimulation, bladder neuroprosthesis, and cortical stimulation for restoring sensation. We also discuss neural recording technologies that may provide command or feedback signals for neuroprosthetic devices.

CONCLUSION/CLINICAL RELEVANCE

Neuroprostheses have begun to address the priorities of individuals with SCI, although there remains room for improvement. In addition to continued technological improvements, closing the loop between the technology and the user may help provide intuitive device control with high levels of performance.

Long-latency muscle activity reflects continuous, delayed sensorimotor feedback of task-level and not joint-level error

In both the upper and lower limbs, evidence suggests that short-latency electromyographic (EMG) responses to mechanical perturbations are modulated based on muscle stretch or joint motion, whereas long-latency responses are modulated based on attainment of task-level goals, e.g., desired direction of limb movement. We hypothesized that long-latency responses are modulated continuously by task-level error feedback. Previously, we identified an error-based sensorimotor feedback transformation that describes the time course of EMG responses to ramp-and-hold perturbations during standing balance (Safavynia and Ting 2013; Welch and Ting 2008, 2009). Here, our goals were 1) to test the robustness of the sensorimotor transformation over a richer set of perturbation conditions and postural states; and 2) to explicitly test whether the sensorimotor transformation is based on task-level vs. joint-level error. We developed novel perturbation trains of acceleration pulses such that perturbations were applied when the body deviated from the desired, upright state while recovering from preceding perturbations. The entire time course of EMG responses (∼4 s) in an antagonistic muscle pair was reconstructed using a weighted sum of center of mass (CoM) kinematics preceding EMGs at long-latency delays (∼100 ms). Furthermore, CoM and joint kinematic trajectories became decorrelated during perturbation trains, allowing us to explicitly compare task-level vs. joint feedback in the same experimental condition. Reconstruction of EMGs was poorer using joint kinematics compared with CoM kinematics and required unphysiologically short (∼10 ms) delays. Thus continuous, long-latency feedback of task-level variables may be a common mechanism regulating long-latency responses in the upper and lower limbs.

Real-time control of hind limb functional electrical stimulation using feedback from dorsal root ganglia recordings

OBJECTIVE

Functional electrical stimulation (FES) approaches often utilize an open-loop controller to drive state transitions. The addition of sensory feedback may allow for closed-loop control that can respond effectively to perturbations and muscle fatigue.

APPROACH

We evaluated the use of natural sensory nerve signals obtained with penetrating microelectrode arrays in lumbar dorsal root ganglia (DRG) as real-time feedback for closed-loop control of FES-generated hind limb stepping in anesthetized cats.

MAIN RESULTS

Leg position feedback was obtained in near real-time at 50 ms intervals by decoding the firing rates of more than 120 DRG neurons recorded simultaneously. Over 5 m of effective linear distance was traversed during closed-loop stepping trials in each of two cats. The controller compensated effectively for perturbations in the stepping path when DRG sensory feedback was provided. The presence of stimulation artifacts and the quality of DRG unit sorting did not significantly affect the accuracy of leg position feedback obtained from the linear decoding model as long as at least 20 DRG units were included in the model.

SIGNIFICANCE

This work demonstrates the feasibility and utility of closed-loop FES control based on natural neural sensors. Further work is needed to improve the controller and electrode technologies and to evaluate long-term viability.

Population coding of forelimb joint kinematics by peripheral afferents in monkeys

Various peripheral receptors provide information concerning position and movement to the central nervous system to achieve complex and dexterous movements of forelimbs in primates. The response properties of single afferent receptors to movements at a single joint have been examined in detail, but the population coding of peripheral afferents remains poorly defined. In this study, we obtained multichannel recordings from dorsal root ganglion (DRG) neurons in cervical segments of monkeys. We applied the sparse linear regression (SLiR) algorithm to the recordings, which selects useful input signals to reconstruct movement kinematics. Multichannel recordings of peripheral afferents were performed by inserting multi-electrode arrays into the DRGs of lower cervical segments in two anesthetized monkeys. A total of 112 and 92 units were responsive to the passive joint movements or the skin stimulation with a painting brush in Monkey 1 and Monkey 2, respectively. Using the SLiR algorithm, we reconstructed the temporal changes of joint angle, angular velocity, and acceleration at the elbow, wrist, and finger joints from temporal firing patterns of the DRG neurons. By automatically selecting a subset of recorded units, the SLiR achieved superior generalization performance compared with a regularized linear regression algorithm. The SLiR selected not only putative muscle units that were responsive to only the passive movements, but also a number of putative cutaneous units responsive to the skin stimulation. These results suggested that an ensemble of peripheral primary afferents that contains both putative muscle and cutaneous units encode forelimb joint kinematics of non-human primates.

In vivo effects of L1 coating on inflammation and neuronal health at the electrode-tissue interface in rat spinal cord and dorsal root ganglion

The spinal cord (SC) and dorsal root ganglion (DRG) are target implantation regions for neural prosthetics, but the tissue-electrode interface in these regions is not well-studied. To improve understanding of these locations, the tissue reactions around implanted electrodes were characterized. L1, an adhesion molecule shown to maintain neuronal density and reduce gliosis in brain tissue, was then evaluated in SC and DRG implants. Following L1 immobilization onto neural electrodes, the bioactivities of the coatings were verified in vitro using neuron, astrocyte and microglia cultures. Non-modified and L1-coated electrodes were implanted into adult rats for 1 or 4 weeks. Hematoxylin and eosin staining along with cell-type specific antibodies were used to characterize the tissue response. In the SC and DRG, cells aggregated at the electrode-tissue interface. Microglia staining was more intense around the implant site and decreased with distance from the interface. Neurofilament staining in both locations decreased or was absent around the implant, compared with surrounding tissue. With L1, neurofilament staining was significantly increased while neuronal cell death decreased. These results indicate that L1-modified electrodes may result in an improved chronic neural interface and will be evaluated in recording and stimulation studies.

Stationary wavelet transform and higher order statistical analyses of intrafascicular nerve recordings

Nerve signals were recorded from the sciatic nerve of the rabbits in the acute experiments with multi-channel thin-film longitudinal intrafascicular electrodes. 5.5 s sequences of quiescent and high-level nerve activity were spectrally decomposed by applying a ten-level stationary wavelet transform with the Daubechies 10 (Db10) mother wavelet. Then, the statistical distributions of the raw and subband-decomposed sequences were estimated and used to fit a fourth-order Pearson distribution as well as check for normality. The results indicated that the raw and decomposed background and high-level nerve activity distributions were nearly zero-mean and non-skew. All distributions with the frequency content above 187.5 Hz were leptokurtic except for the first-level decomposition representing frequencies in the subband between 12 and 24 kHz, which was Gaussian. This suggests that nerve activity acts to change the statistical distribution of the recording. The results further demonstrated that quiescent recording contained a mixture of an underlying pink noise and low-level nerve activity that could not be silenced. The signal-to-noise ratios based upon the standard deviation (SD) and kurtosis were estimated, and the latter was found as an effective measure for monitoring the nerve activity residing in different frequency subbands. The nerve activity modulated kurtosis along with SD, suggesting that the joint use of SD and kurtosis could improve the stability and detection accuracy of spike-detection algorithms. Finally, synthesizing the reconstructed subband signals following denoising based upon the higher order statistics of the subband-decomposed coefficient sequences allowed us to effectively purify the signal without distorting spike shape.

Online feedback control of functional electrical stimulation using dorsal root ganglia recordings

In neuroprostheses that use functional electrical stimulation (FES) to restore motor function, closed-loop feedback control may compensate for muscle fatigue, perturbations and nonlinearities in the behavior of the effected muscles. Kinematic state information is naturally represented in the firing rates of primary afferent neurons, which may be recorded with multi-electrode arrays at the level of the dorsal root ganglia (DRG). Previous work in cats has shown that it is feasible to estimate the kinematic state of the hind limb with a multivariate linear regression model of the neural activity in the DRG. In this study we extend these results to estimate the limb state in real-time during intramuscular stimulation in an anesthetized cat. Furthermore, we used the limb state estimates as feedback to a finite state FES controller to generate rudimentary walking behavior. This work demonstrates the feasibility of using DRG activity in a closed-loop FES system.

Single- and multi-unit activity recorded from the surface of the dorsal root ganglia with non-penetrating electrode arrays

Non-penetrating surface electrode recording techniques are typically associated with field potential recordings, while extracellular recordings from single neurons are made using penetrating metal wire or microfabricated microelectrode arrays. Here, we report on single- and multi-unit neuronal recordings made using non-penetrating electrodes placed on the epineural surface of the dorsal root ganglia (DRG). Across four experiments in anesthetized cats, approximately 40% of the electrodes recorded single- and multi-unit spiking activity with spike-rates that covaried significantly with hindlimb movement. In two intraoperative experiments in humans, compound activity was recorded from the DRG surface in response to peripheral stimulation of the common peroneal nerve. This approach may have advantages over penetrating electrode arrays in terms of clinical acceptability and recording longevity.

Estimating bladder pressure from sacral dorsal root ganglia recordings

Individuals with dysfunctional bladders may benefit from devices that track the bladder state. Recordings from pelvic and sacral nerve cuffs can detect bladder contractions, however they often have low signal quality and are susceptible to interference from non-bladder signals. Microelectrode recordings from sacral dorsal root ganglia (DRG) neurons may provide an alternate source for obtaining high quality sensory signals for bladder pressure monitoring. In this study, penetrating microelectrode arrays were inserted in the S1 and S2 DRG in two cats to record afferent spiking activity at different bladder pressures. Multivariate linear regression models were used to estimate bladder pressure from the spiking activity of DRG neurons. The best estimates were obtained with populations of 5-10 units primarily from the S2 DRG, with root mean square errors of 3.0-3.2 cm H(2)O (correlation coefficients of 0.5-0.9). This work demonstrates the feasibility of monitoring bladder pressure from DRG recordings.

Restoring stepping after spinal cord injury using intraspinal microstimulation and novel control strategies

The overall objective of this project is to develop a feedback-driven intraspinal microstimulation (ISMS) system. We hypothesize that ISMS will enhance the functionality of stepping by reducing muscle fatigue and producing synergistic movements by activating neural networks in the spinal cord. In the present pilot study, the controller was tested with ISMS and external sensors (force plates, gyroscopes, and accelerometers). Cats were partially supported in a sling and bi-laterally stepped overground on a 4-m instrumented walkway. The walkway had variable friction. Limb angle was controlled to within 10° even in the presence of variable friction. Peak ground reaction forces in each limb were approximately 12% of body weight (12.5% was full load bearing in this experimental setup); rarely, the total supportive force briefly decreased to as low as 4.1%. Magnetic resonance images were acquired of the excised spinal cord and the implanted array. The majority of electrodes (75%) were implanted successfully into their target regions. This represents the first successful application of ISMS for overground walking.

Optimization of muscle activity for task-level goals predicts complex changes in limb forces across biomechanical contexts

Optimality principles have been proposed as a general framework for understanding motor control in animals and humans largely based on their ability to predict general features movement in idealized motor tasks. However, generalizing these concepts past proof-of-principle to understand the neuromechanical transformation from task-level control to detailed execution-level muscle activity and forces during behaviorally-relevant motor tasks has proved difficult. In an unrestrained balance task in cats, we demonstrate that achieving task-level constraints center of mass forces and moments while minimizing control effort predicts detailed patterns of muscle activity and ground reaction forces in an anatomically-realistic musculoskeletal model. Whereas optimization is typically used to resolve redundancy at a single level of the motor hierarchy, we simultaneously resolved redundancy across both muscles and limbs and directly compared predictions to experimental measures across multiple perturbation directions that elicit different intra- and interlimb coordination patterns. Further, although some candidate task-level variables and cost functions generated indistinguishable predictions in a single biomechanical context, we identified a common optimization framework that could predict up to 48 experimental conditions per animal (n = 3) across both perturbation directions and different biomechanical contexts created by altering animals' postural configuration. Predictions were further improved by imposing experimentally-derived muscle synergy constraints, suggesting additional task variables or costs that may be relevant to the neural control of balance. These results suggested that reduced-dimension neural control mechanisms such as muscle synergies can achieve similar kinetics to the optimal solution, but with increased control effort (≈2×) compared to individual muscle control. Our results are consistent with the idea that hierarchical, task-level neural control mechanisms previously associated with voluntary tasks may also be used in automatic brainstem-mediated pathways for balance.

Integrating rehabilitation engineering technology with biologics

Rehabilitation engineers apply engineering principles to improve function or to solve challenges faced by persons with disabilities. It is critical to integrate the knowledge of biologics into the process of rehabilitation engineering to advance the field and maximize potential benefits to patients. Some applications in particular demonstrate the value of a symbiotic relationship between biologics and rehabilitation engineering. In this review we illustrate how researchers working with neural interfaces and integrated prosthetics, assistive technology, and biologics data collection are currently integrating these 2 fields. We also discuss the potential for further integration of biologics and rehabilitation engineering to deliver the best technologies and treatments to patients. Engineers and clinicians must work together to develop technologies that meet clinical needs and are accessible to the intended patient population.

Modality-specific axonal regeneration: toward selective regenerative neural interfaces

Regenerative peripheral nerve interfaces have been proposed as viable alternatives for the natural control of robotic prosthetic devices. However, sensory and motor axons at the neural interface are of mixed sub-modality types, which difficult the specific recording from motor axons and the eliciting of precise sensory modalities through selective stimulation. Here we evaluated the possibility of using type specific neurotrophins to preferentially entice the regeneration of defined axonal populations from transected peripheral nerves into separate compartments. Segregation of mixed sensory fibers from dorsal root ganglion neurons was evaluated in vitro by compartmentalized diffusion delivery of nerve growth factor (NGF) and neurotrophin-3 (NT-3), to preferentially entice the growth of TrkA+ nociceptive and TrkC+ proprioceptive subsets of sensory neurons, respectively. The average axon length in the NGF channel increased 2.5-fold compared to that in saline or NT-3, whereas the number of branches increased threefold in the NT-3 channels. These results were confirmed using a 3D “Y”-shaped in vitro assay showing that the arm containing NGF was able to entice a fivefold increase in axonal length of unbranched fibers. To address if such segregation can be enticed in vivo, a “Y”-shaped tubing was used to allow regeneration of the transected adult rat sciatic nerve into separate compartments filled with either NFG or NT-3. A significant increase in the number of CGRP+ pain fibers were attracted toward the sural nerve, while N-52+ large-diameter axons were observed in the tibial and NT-3 compartments. This study demonstrates the guided enrichment of sensory axons in specific regenerative chambers, and supports the notion that neurotrophic factors can be used to segregate sensory and perhaps motor axons in separate peripheral interfaces.

Task-level feedback can explain temporal recruitment of spatially fixed muscle synergies throughout postural perturbations

Recent evidence suggests that complex spatiotemporal patterns of muscle activity can be explained with a low-dimensional set of muscle synergies or M-modes. While it is clear that both spatial and temporal aspects of muscle coordination may be low dimensional, constraints on spatial versus temporal features of muscle coordination likely involve different neural control mechanisms. We hypothesized that the low-dimensional spatial and temporal features of muscle coordination are independent of each other. We further hypothesized that in reactive feedback tasks, spatially fixed muscle coordination patterns-or muscle synergies-are hierarchically recruited via time-varying neural commands based on delayed task-level feedback. We explicitly compared the ability of spatially fixed (SF) versus temporally fixed (TF) muscle synergies to reconstruct the entire time course of muscle activity during postural responses to anterior-posterior support-surface translations. While both SF and TF muscle synergies could account for EMG variability in a postural task, SF muscle synergies produced more consistent and physiologically interpretable results than TF muscle synergies during postural responses to perturbations. Moreover, a majority of SF muscle synergies were consistent in structure when extracted from epochs throughout postural responses. Temporal patterns of SF muscle synergy recruitment were well-reconstructed by delayed feedback of center of mass (CoM) kinematics and reproduced EMG activity of multiple muscles. Consistent with the idea that independent and hierarchical low-dimensional neural control structures define spatial and temporal patterns of muscle activity, our results suggest that CoM kinematics are a task variable used to recruit SF muscle synergies for feedback control of balance.

Limb-state information encoded by peripheral and central somatosensory neurons: implications for an afferent interface

A major issue to be addressed in the development of neural interfaces for prosthetic control is the need for somatosensory feedback. Here, we investigate two possible strategies: electrical stimulation of either dorsal root ganglia (DRG) or primary somatosensory cortex (S1). In each approach, we must determine a model that reflects the representation of limb state in terms of neural discharge. This model can then be used to design stimuli that artificially activate the nervous system to convey information about limb state to the subject. Electrically activating DRG neurons using naturalistic stimulus patterns, modeled on recordings made during passive limb movement, evoked activity in S1 that was similar to that of the original movement. We also found that S1 neural populations could accurately discriminate different patterns of DRG stimulation across a wide range of stimulus pulse-rates. In studying the neural coding in S1, we also decoded the kinematics of active limb movement using multi-electrode recordings in the monkey. Neurons having both proprioceptive and cutaneous receptive fields contributed equally to this decoding. Some neurons were most informative of limb state in the recent past, but many others appeared to signal upcoming movements suggesting that they also were modulated by an efference copy signal. Finally, we show that a monkey was able to detect stimulation through a large percentage of electrodes implanted in area 2. We discuss the design of appropriate stimulus paradigms for conveying time-varying limb state information, and the relative merits and limitations of central and peripheral approaches.

Multielectrode array recordings of bladder and perineal primary afferent activity from the sacral dorsal root ganglia

The development of bladder and bowel neuroprostheses may benefit from the use of sensory feedback. We evaluated the use of high-density penetrating microelectrode arrays in sacral dorsal root ganglia (DRG) for recording bladder and perineal afferent activity. Arrays were inserted in S1 and S2 DRG in three anesthetized cats. Neural signals were recorded while the bladder volume was modulated and mechanical stimuli were applied to the perineal region. In two experiments, 48 units were observed that tracked bladder pressure with their firing rates (79% from S2). At least 50 additional units in each of the three experiments (274 total; 60% from S2) had a significant change in their firing rates during one or more perineal stimulation trials. This study shows the feasibility of obtaining bladder-state information and other feedback signals from the pelvic region with a sacral DRG electrode interface located in a single level. This natural source of feedback would be valuable for providing closed-loop control of bladder or other pelvic neuroprostheses.