Formation of a novel supraspinal-spinal connectome that relearns the same motor task after complete paralysis.

Having observed that electrical spinal cord stimulation and training enabled four patients with paraplegia with motor complete paralysis to regain voluntary leg movement, the underlying mechanisms involved in forming the newly established supraspinal-spinal functional connectivity have become of great interest. van den Brand et al. (Science 336: 1182-1185, 2012) subsequently, demonstrated the recovery, in response to spinal electro-neuromodulation and locomotor training, of voluntary stepping of the lower limbs in rats that received a lesion that is assumed to eliminate all long-descending cortical axons that project to lumbosacral segments. Here, we used a similar spinal lesion in rats to eliminate long-descending axons to determine whether a novel, trained motor behavior triggered by a unique auditory cue learned before a spinal lesion, could recover after the lesion. Hindlimb stepping recovered 1 mo after the spinal injury, but only after 2 mo, the novel and unique audio-triggered behavior was recovered, meaning that not only was a novel connectivity formed but also further evidence suggested that this highly unique behavioral response was independent of the recovery of the circuitry that generated stepping. The unique features of the newly formed supraspinal-spinal connections that mediated the recovery of the trained behavior is consistent with a guidance mechanism(s) that are highly use dependent.NEW & NOTEWORTHY Electrical spinal cord stimulation has enabled patients with paraplegia to regain voluntary leg movement, and so the underlying mechanisms involved in this recovery are of great interest. Here, we demonstrate in rodents the recovery of trained motor behavior after a spinal lesion. Rodents were trained to kick their right hindlimb in response to an auditory cue. This behavior recovered 2 mo after the paralyzing spinal cord injury but only with the assistance of electrical spinal cord stimulation.

Motor Control After Human SCI Through Activation of Muscle Synergies Under Spinal Cord Stimulation.

Spinal cord stimulation (SCS) has enabled motor recovery in paraplegics with motor complete spinal cord injury (SCI). However, the physiological mechanisms underlying this recovery are unknown. This paper analyzes muscle synergies in two motor complete SCI patients under SCS during standing and compares them with muscle synergies in healthy subjects, in order to help elucidate the mechanisms that enable motor control through SCS. One challenge is that standard muscle synergy extraction algorithms, such as non-negative matrix factorization (NMF), fail when applied to SCI patients under SCS. We develop a new algorithm-rShiftNMF-to extract muscle synergies in these cases. We find muscle synergies extracted by rShiftNMF are significantly better at interpreting electromyography (EMG) activity, and resulting synergy features are more physiologically meaningful. By analyzing muscle synergies from SCI patients and healthy subjects, we find that: 1) SCI patients rely significantly on muscle synergy activation to generate motor activity; 2) interleaving SCS can selectively activate an additional muscle synergy that is critical to SCI standing; and 3) muscle synergies extracted from SCI patients under SCS differ substantially from those extracted from healthy subjects. We provide evidence that after spinal cord injury, SCS influences motor function through muscle synergy activation.

Self-Assisted Standing Enabled by Non-Invasive Spinal Stimulation after Spinal Cord Injury.

Neuromodulation of spinal networks can improve motor control after spinal cord injury (SCI). The objectives of this study were to (1) determine whether individuals with chronic paralysis can stand with the aid of non-invasive electrical spinal stimulation with their knees and hips extended without trainer assistance, and (2) investigate whether postural control can be further improved following repeated sessions of stand training. Using a double-blind, balanced, within-subject cross-over, and sham-controlled study design, 15 individuals with SCI of various severity received transcutaneous electrical spinal stimulation to regain self-assisted standing. The primary outcomes included qualitative comparison of need of external assistance for knee and hip extension provided by trainers during standing without and in the presence of stimulation in the same participants, as well as quantitative measures, such as the level of knee assistance and amount of time spent standing without trainer assistance. None of the participants could stand unassisted without stimulation or in the presence of sham stimulation. With stimulation all participants could maintain upright standing with minimum and some (n = 7) without external assistance applied to the knees or hips, using their hands for upper body balance as needed. Quality of balance control was practice-dependent, and improved with subsequent training. During self-initiated body-weight displacements in standing enabled by spinal stimulation, high levels of leg muscle activity emerged, and depended on the amount of muscle loading. Our findings indicate that the lumbosacral spinal networks can be modulated transcutaneously using electrical spinal stimulation to facilitate self-assisted standing after chronic motor and sensory complete paralysis.

Extraction of Muscle Synergies in Spinal Cord Injured Patients.

Muscle synergies encode motor activity as a linear superposition of multiple motor units composed of a temporal command exciting a specific network of muscles. This study examines muscle synergies derived from simple standing studies of a complete spinal cord injury (SCI) patient under epidural spinal stimulation. A popular technique for extracting these synergies from EMG data is non-negative matrix factorization (NNMF). However, standard NNMF algorithms do not allow for physiological delays for a neural signal to reach different muscles. These delays are prevalent in SCI patients under spinal stimulation, and so we propose a new algorithm (regularized ShiftNMF) to extract muscle synergies which account for signal delays. We find muscle synergies extracted by the regularized ShiftNMF algorithm are significantly better at reconstructing EMG activity, and the resulting features are physiologically consistent and more useful in describing patient behavior.

An Active Learning Algorithm for Control of Epidural Electrostimulation.

Epidural electrostimulation has shown promise for spinal cord injury therapy. However, finding effective stimuli on the multi-electrode stimulating arrays employed requires a laborious manual search of a vast space for each patient. Widespread clinical application of these techniques would be greatly facilitated by an autonomous, algorithmic system which choses stimuli to simultaneously deliver effective therapy and explore this space. We propose a method based on GP-BUCB, a Gaussian process bandit algorithm. In n = 4 spinally transected rats, we implant epidural electrode arrays and examine the algorithm's performance in selecting bipolar stimuli to elicit specified muscle responses. These responses are compared with temporally interleaved intra-animal stimulus selections by a human expert. GP-BUCB successfully controlled the spinal electrostimulation preparation in 37 testing sessions, selecting 670 stimuli. These sessions included sustained autonomous operations (ten-session duration). Delivered performance with respect to the specified metric was as good as or better than that of the human expert. Despite receiving no information as to anatomically likely locations of effective stimuli, GP-BUCB also consistently discovered such a pattern. Further, GP-BUCB was able to extrapolate from previous sessions' results to make predictions about performance in new testing sessions, while remaining sufficiently flexible to capture temporal variability. These results provide validation for applying automated stimulus selection methods to the problem of spinal cord injury therapy.

Expert-like performance of an autonomous spike tracking algorithm in isolating and maintaining single units in the macaque cortex.

Isolating action potentials of a single neuron (unit) is essential for intra-cortical neurophysiological recordings. Yet, during extracellular recordings in semi-chronic awake preparations, the relationship between neuronal soma and the recording electrode is typically not stationary. Neuronal waveforms often change in shape, and in the absence of counter-measures, merge with the background noise. To avoid this, experimenters can repeatedly re-adjust electrode positions to maintain the shapes of isolated spikes. In recordings with a larger number of electrodes, this process becomes extremely difficult. We report the performance of an automated algorithm that tracks neurons to obtain well isolated spiking, and autonomously adjusts electrode position to maintain good isolation. We tested the performance of this algorithm in isolating units with multiple individually adjustable micro-electrodes in a cortical surface area of macaque monkeys. We compared the performance in terms of signal quality and signal stability against passive placement of microelectrodes and against the performance of three human experts. The results show that our SpikeTrack2 algorithm achieves significantly better signal quality compared to passive placement. It is as least as good as humans in initially finding and isolating units, and better as the average and at least as good as the most proficient of three human experimenters in maintaining signal quality and signal stability. The autonomous tracking performance, the scalability of the system to large numbers of individual channels, and the possibility to objectify single unit recording criteria makes SpikeTrack2 a highly valuable tool for all multi-channel recording systems with individually adjustable electrodes.

A Bayesian clustering method for tracking neural signals over successive intervals.

This paper introduces a new, unsupervised method for sorting and tracking the action potentials of individual neurons in multiunit extracellular recordings. Presuming the data are divided into short, sequential recording intervals, the core of our strategy relies upon an extension of a traditional mixture model approach that incorporates clustering results from the preceding interval in a Bayesian manner, while still allowing for signal nonstationarity and changing numbers of recorded neurons. As a natural byproduct of the sorting method, current and prior signal clusters can be matched over time in order to track persisting neurons. We also develop techniques to use prior data to appropriately seed the clustering algorithm and select the model class. We present results in a principal components space; however, the algorithm may be applied in any feature space where the distribution of a neuron's spikes may be modeled as Gaussian. Applications of this signal classification method to recordings from macaque parietal cortex show that it provides significantly more consistent clustering and tracking results than traditional methods based on expectation-maximization optimization of mixture models. This consistent tracking ability is crucial for intended applications of the method.

Wing and body motion during flight initiation in Drosophila revealed by automated visual tracking.

The fruit fly Drosophila melanogaster is a widely used model organism in studies of genetics, developmental biology and biomechanics. One limitation for exploiting Drosophila as a model system for behavioral neurobiology is that measuring body kinematics during behavior is labor intensive and subjective. In order to quantify flight kinematics during different types of maneuvers, we have developed a visual tracking system that estimates the posture of the fly from multiple calibrated cameras. An accurate geometric fly model is designed using unit quaternions to capture complex body and wing rotations, which are automatically fitted to the images in each time frame. Our approach works across a range of flight behaviors, while also being robust to common environmental clutter. The tracking system is used in this paper to compare wing and body motion during both voluntary and escape take-offs. Using our automated algorithms, we are able to measure stroke amplitude, geometric angle of attack and other parameters important to a mechanistic understanding of flapping flight. When compared with manual tracking methods, the algorithm estimates body position within 4.4+/-1.3% of the body length, while body orientation is measured within 6.5+/-1.9 deg. (roll), 3.2+/-1.3 deg. (pitch) and 3.4+/-1.6 deg. (yaw) on average across six videos. Similarly, stroke amplitude and deviation are estimated within 3.3 deg. and 2.1 deg., while angle of attack is typically measured within 8.8 deg. comparing against a human digitizer. Using our automated tracker, we analyzed a total of eight voluntary and two escape take-offs. These sequences show that Drosophila melanogaster do not utilize clap and fling during take-off and are able to modify their wing kinematics from one wingstroke to the next. Our approach should enable biomechanists and ethologists to process much larger datasets than possible at present and, therefore, accelerate insight into the mechanisms of free-flight maneuvers of flying insects.

Automated visual tracking for studying the ontogeny of zebrafish swimming.

The zebrafish Danio rerio is a widely used model organism in studies of genetics, developmental biology, and recently, biomechanics. In order to quantify changes in swimming during all stages of development, we have developed a visual tracking system that estimates the posture of fish. Our current approach assumes planar motion of the fish, given image sequences taken from a top view. An accurate geometric fish model is automatically designed and fit to the images at each time frame. Our approach works across a range of fish shapes and sizes and is therefore well suited for studying the ontogeny of fish swimming, while also being robust to common environmental occlusions. Our current analysis focuses on measuring the influence of vertebra development on the swimming capabilities of zebrafish. We examine wild-type zebrafish and mutants with stiff vertebrae (stocksteif) and quantify their body kinematics as a function of their development from larvae to adult (mutants made available by the Hubrecht laboratory, The Netherlands). By tracking the fish, we are able to measure the curvature and net acceleration along the body that result from the fish's body wave. Here, we demonstrate the capabilities of the tracking system for the escape response of wild-type zebrafish and stocksteif mutant zebrafish. The response was filmed with a digital high-speed camera at 1500 frames s(-1). Our approach enables biomechanists and ethologists to process much larger datasets than possible at present. Our automated tracking scheme can therefore accelerate insight in the swimming behavior of many species of (developing) fish.

Training locomotor networks.

For a complete adult spinal rat to regain some weight-bearing stepping capability, it appears that a sequence of specific proprioceptive inputs that are similar, but not identical, from step to step must be generated over repetitive step cycles. Furthermore, these cycles must include the activation of specific neural circuits that are intrinsic to the lumbosacral spinal cord segments. For these sensorimotor pathways to be effective in generating stepping, the spinal circuitry must be modulated to an appropriate excitability level. This level of modulation is sustained from supraspinal input in intact, but not spinal, rats. In a series of experiments with complete spinal rats, we have shown that an appropriate level of excitability of the spinal circuitry can be achieved using widely different means. For example, this modulation level can be acquired pharmacologically, via epidural electrical stimulation over specific lumbosacral spinal cord segments, and/or by use-dependent mechanisms such as step or stand training. Evidence as to how each of these treatments can "tune" the spinal circuitry to a "physiological state" that enables it to respond appropriately to proprioceptive input will be presented. We have found that each of these interventions can enable the proprioceptive input to actually control extensive details that define the dynamics of stepping over a range of speeds, loads, and directions. A series of experiments will be described that illustrate sensory control of stepping and standing after a spinal cord injury and the necessity for the "physiological state" of the spinal circuitry to be modulated within a critical window of excitability for this control to be manifested. The present findings have important consequences not only for our understanding of how the motor pattern for stepping is formed, but also for the design of rehabilitation intervention to restore lumbosacral circuit function in humans following a spinal cord injury.

Omnidirectional sensory and motor volumes in electric fish.

Active sensing organisms, such as bats, dolphins, and weakly electric fish, generate a 3-D space for active sensation by emitting self-generated energy into the environment. For a weakly electric fish, we demonstrate that the electrosensory space for prey detection has an unusual, omnidirectional shape. We compare this sensory volume with the animal's motor volume--the volume swept out by the body over selected time intervals and over the time it takes to come to a stop from typical hunting velocities. We find that the motor volume has a similar omnidirectional shape, which can be attributed to the fish's backward-swimming capabilities and body dynamics. We assessed the electrosensory space for prey detection by analyzing simulated changes in spiking activity of primary electrosensory afferents during empirically measured and synthetic prey capture trials. The animal's motor volume was reconstructed from video recordings of body motion during prey capture behavior. Our results suggest that in weakly electric fish, there is a close connection between the shape of the sensory and motor volumes. We consider three general spatial relationships between 3-D sensory and motor volumes in active and passive-sensing animals, and we examine hypotheses about these relationships in the context of the volumes we quantify for weakly electric fish. We propose that the ratio of the sensory volume to the motor volume provides insight into behavioral control strategies across all animals.

Implications of assist-as-needed robotic step training after a complete spinal cord injury on intrinsic strategies of motor learning.

Robotic training paradigms that enforce a fixed kinematic control might be suboptimal for rehabilitative training because they abolish variability, an intrinsic property of neuromuscular control (Jezernik et al., 2003). In the present study we introduce "assist-as-needed" (AAN) robotic training paradigms for rehabilitation of spinal cord injury subjects. To test the efficacy of these robotic control strategies to teach spinal mice to step, we divided 27 adult female Swiss-Webster mice randomly into three groups. Each group was trained robotically by using one of three control strategies: a fixed training trajectory (Fixed group), an AAN training paradigm without interlimb coordination (Band group), and an AAN training paradigm with bilateral hindlimb coordination (Window group). Beginning at 14 d after a complete midthoracic spinal cord transection, the mice were trained daily (10 min/d, 5 d/week) to step on a treadmill 10 min after the administration of quipazine (0.5 mg/kg), a serotonin agonist, for a period of 6 weeks. During weekly performance evaluations, the mice trained with the AAN window paradigm generally showed the highest level of recovery as measured by the number, consistency, and periodicity of steps during the testing sessions. In all three measurements there were no significant differences between the Band and the Fixed training groups. These results indicate that the window training approach, which includes loose alternating interlimb coordination, is more effective than a fixed trajectory paradigm with rigid alternating interlimb coordination or an AAN paradigm without any interlimb constraints in promoting robust postinjury stepping behavior.

A control algorithm for autonomous optimization of extracellular recordings.

This paper develops a control algorithm that can autonomously position an electrode so as to find and then maintain an optimal extracellular recording position. The algorithm was developed and tested in a two-neuron computational model representative of the cells found in cerebral cortex. The algorithm is based on a stochastic optimization of a suitably defined signal quality metric and is shown capable of finding the optimal recording position along representative sampling directions, as well as maintaining the optimal signal quality in the face of modeled tissue movements. The application of the algorithm to acute neurophysiological recording experiments and its potential implications to chronic recording electrode arrays are discussed.

Spinal cord-transected mice learn to step in response to quipazine treatment and robotic training.

In the present study, concurrent treatment with robotic step training and a serotonin agonist, quipazine, generated significant recovery of locomotor function in complete spinal cord-transected mice (T7-T9) that otherwise could not step. The extent of recovery achieved when these treatments were combined exceeded that obtained when either treatment was applied independently. We quantitatively analyzed the stepping characteristics of spinal mice after alternatively administering no training, manual training, robotic training, quipazine treatment, or a combination of robotic training with quipazine treatment, to examine the mechanisms by which training and quipazine treatment promote functional recovery. Using fast Fourier transform and principal components analysis, significant improvements in the step rhythm, step shape consistency, and number of weight-bearing steps were observed in robotically trained compared with manually trained or nontrained mice. In contrast, manual training had no effect on stepping performance, yielding no improvement compared with nontrained mice. Daily bolus quipazine treatment acutely improved the step shape consistency and number of steps executed by both robotically trained and nontrained mice, but these improvements did not persist after quipazine was withdrawn. At the dosage used (0.5 mg/kg body weight), quipazine appeared to facilitate, rather than directly generate, stepping, by enabling the spinal cord neural circuitry to process specific patterns of sensory information associated with weight-bearing stepping. Via this mechanism, quipazine treatment enhanced kinematically appropriate robotic training. When administered intermittently during an extended period of robotic training, quipazine revealed training-induced stepping improvements that were masked in the absence of the pharmacological treatment.

Spike detection using the continuous wavelet transform.

This paper combines wavelet transforms with basic detection theory to develop a new unsupervised method for robustly detecting and localizing spikes in noisy neural recordings. The method does not require the construction of templates, or the supervised setting of thresholds. We present extensive Monte Carlo simulations, based on actual extracellular recordings, to show that this technique surpasses other commonly used methods in a wide variety of recording conditions. We further demonstrate that falsely detected spikes corresponding to our method resemble actual spikes more than the false positives of other techniques such as amplitude thresholding. Moreover, the simplicity of the method allows for nearly real-time execution.

Semi-chronic motorized microdrive and control algorithm for autonomously isolating and maintaining optimal extracellular action potentials.

A system was developed that can autonomously position recording electrodes to isolate and maintain optimal quality extracellular signals. The system consists of a novel motorized miniature recording microdrive and a control algorithm. The microdrive was designed for chronic operation and can independently position four glass-coated Pt-Ir electrodes with micrometer precision over a 5-mm range using small (3 mm diam) piezoelectric linear actuators. The autonomous positioning algorithm is designed to detect, align, and cluster action potentials and then command the microdrive to optimize and maintain the neural signal. This system is shown to be capable of autonomous operation in monkey cortical tissue.

Neural prosthetic control signals from plan activity.

The prospect of assisting disabled patients by translating neural activity from the brain into control signals for prosthetic devices, has flourished in recent years. Current systems rely on neural activity present during natural arm movements. We propose here that neural activity present before or even without natural arm movements can provide an important, and potentially advantageous, source of control signals. To demonstrate how control signals can be derived from such plan activity we performed a computational study with neural activity previously recorded from the posterior parietal cortex of rhesus monkeys planning arm movements. We employed maximum likelihood decoders to estimate movement direction and to drive finite state machines governing when to move. Performance exceeded 90% with as few as 40 neurons.