Brian Hillen

Accelerating locomotor recovery after incomplete spinal injury

A traumatic spinal injury can destroy cells, irreparably damage axons, and trigger a cascade of biochemical responses that increase the extent of injury. Although damaged central nervous system axons do not regrow well naturally, the distributed nature of the nervous system and its capacity to adapt provide opportunities for recovery of function. It is apparent that activity‐dependent plasticity plays a role in this recovery and that the endogenous response to injury heightens the capacity for recovery for at least several weeks postinjury. To restore locomotor function, researchers have investigated the use of treadmill‐based training, robots, and electrical stimulation to tap into adaptive activity‐dependent processes. The current challenge is to maximize the degree of functional recovery. This manuscript reviews the endogenous neural system response to injury, and reviews data and presents novel analyses of these from a rat model of contusion injury that demonstrates how a targeted intervention can accelerate recovery, presumably by engaging processes that underlie activity‐dependent plasticity.

Joint-specific changes in locomotor complexity in the absence of muscle atrophy following incomplete spinal cord injury.

BackgroundFollowing incomplete spinal cord injury (iSCI), descending drive is impaired, possibly leading to a decrease in the complexity of gait. To test the hypothesis that iSCI impairs gait coordination and decreases locomotor complexity, we collected 3D joint angle kinematics and muscle parameters of rats with a sham or an incomplete spinal cord injury.Methods12 adult, female, Long-Evans rats, 6 sham and 6 mild-moderate T8 iSCI, were tested 4 weeks following injury. The Basso Beattie Bresnahan locomotor score was used to verify injury severity. Animals had reflective markers placed on the bony prominences of their limb joints and were filmed in 3D while walking on a treadmill. Joint angles and segment motion were analyzed quantitatively, and complexity of joint angle trajectory and overall gait were calculated using permutation entropy and principal component analysis, respectively. Following treadmill testing, the animals were euthanized and hindlimb muscles removed. Excised muscles were tested for mass, density, fiber length, pennation angle, and relaxed sarcomere length.ResultsMuscle parameters were similar between groups with no evidence of muscle atrophy. The animals showed overextension of the ankle, which was compensated for by a decreased range of motion at the knee. Left-right coordination was altered, leading to left and right knee movements that are entirely out of phase, with one joint moving while the other is stationary. Movement patterns remained symmetric. Permutation entropy measures indicated changes in complexity on a joint specific basis, with the largest changes at the ankle. No significant difference was seen using principal component analysis. Rats were able to achieve stable weight bearing locomotion at reasonable speeds on the treadmill despite these deficiencies.ConclusionsDecrease in supraspinal control following iSCI causes a loss of complexity of ankle kinematics. This loss can be entirely due to loss of supraspinal control in the absence of muscle atrophy and may be quantified using permutation entropy. Joint-specific differences in kinematic complexity may be attributed to different sources of motor control. This work indicates the importance of the ankle for rehabilitation interventions following spinal cord injury.

Bio-Inspired Controller on an FPGA Applied to Closed-Loop Diaphragmatic Stimulation.

Cervical spinal cord injury can disrupt connections between the brain respiratory network and the respiratory muscles which can lead to partial or complete loss of ventilatory control and require ventilatory assistance. Unlike current open-loop technology, a closed-loop diaphragmatic pacing system could overcome the drawbacks of manual titration as well as respond to changing ventilation requirements. We present an original bio-inspired assistive technology for real-time ventilation assistance, implemented in a digital configurable Field Programmable Gate Array (FPGA). The bio-inspired controller, which is a spiking neural network (SNN) inspired by the medullary respiratory network, is as robust as a classic controller while having a flexible, low-power and low-cost hardware design. The system was simulated in MATLAB with FPGA-specific constraints and tested with a computational model of rat breathing; the model reproduced experimentally collected respiratory data in eupneic animals. The open-loop version of the bio-inspired controller was implemented on the FPGA. Electrical test bench characterizations confirmed the system functionality. Open and closed-loop paradigm simulations were simulated to test the FPGA system real-time behavior using the rat computational model. The closed-loop system monitors breathing and changes in respiratory demands to drive diaphragmatic stimulation. The simulated results inform future acute animal experiments and constitute the first step toward the development of a neuromorphic, adaptive, compact, low-power, implantable device. The bio-inspired hardware design optimizes the FPGA resource and time costs while harnessing the computational power of spike-based neuromorphic hardware. Its real-time feature makes it suitable for in vivo applications.

Computational model of human ventilation for electrical stimulation following cervical spinal cord injury

Towards the development of an implantable stimulator for respiratory pacing, we have developed a computational model of the human ventilatory system to simulate breathing maneuvers. Following cervical spinal cord injury, control of breathing is lost or diminished. If the loss is severe, patients are dependent on mechanical ventilators or electrical stimulators to regain functional breathing. Existing electrical stimulation based respiratory pacing systems typically activate only the diaphragm and cannot adapt to muscle fatigue, changes in respiratory demands, or changes in electrode quality over time. This model will be used to test stimulation controllers that can adapt to such conditions. The model was developed in Simulink/SimMechanics implementing the physiologically realistic muscle model from MSMS [1,2] using published parameters for muscle geometry [3,4]. Diaphragm geometry was modeled as a pulley system in one dimension. Ventilatory compliance was modeled as a damped spring with non-linear stiffness. Integrated phrenic drive was modeled as either a spaced sawtooth (tidal breathing) or a step function (maximal inspiration). The model was able to reproduce tidal breathing and maximal inspiration for uninjured subjects (see Figure ​Figure1).1). When simulating electrical stimulation without fatigue, maximal inspiration was similar but tidal breathing was higher due to the recruitment of larger motor units first. When simulating electrical stimulation as well as fatigue (50% reduction in maximal force and a decrease in the power of the fast motor units), tidal breathing was similar to the uninjured case (see Figure ​Figure1).1). Thus, even with fatigue, electrically stimulated muscles are strong enough to produce the submaximal contractions needed for tidal breathing. In the future, this model will be enhanced to incorporate additional muscles of respiration as well as atrophy associated with spinal cord injury and will be used to iteratively develop adaptive controllers to achieve functional ventilation and to promote weaning from a mechanical ventilator. Figure 1 Diaphragm displacement during tidal breathing (grey lines) and maximal inspiration (black lines). Uninjured curves (solid lines) reproduce published experimental data. When simulating electrical stimulation (reverse recruitment) and fatigue (a 50% reduction ...

An IC-based controllable stimulator for respiratory muscle stimulation investigations

Functional Electrical Stimulation can be used to restore motor functions loss consecutive to spinal cord injury, such as respiratory deficiency due to paralysis of ventilatory muscles. This paper presents a fully configurable IC-centered stimulator designed to investigate muscle stimulation paradigms. It provides 8 current stimulation channels with high-voltage compliance and real-time operation capabilities, to enable a wide range of FES applications. The stimulator can be used in a standalone mode, or within a closed-loop setup. Primary in vivo results show successful drive of respiratory muscles stimulation using a computer-based dedicated controller.

A versatile fast-development platform applied to closed-loop diaphragmatic pacing

People with cervical spinal cord injury have partial or complete loss of ventilatory control and require ventilator assist. Open-loop diaphragmatic pacing can be utilized to provide this assist. A closed-loop diaphragmatic pacing system could overcome the drawbacks for manual titration of the stimulation and respond to changing ventilatory requirements. We have developed a versatile custom hardware platform dubbed “Multimed” for biosignal acquisition and parallel real-time computation, data display and storage. We have also developed a new rodent model for diaphragmatic pacing. Using these we illustrate, to our knowledge for the first-time, the successful ability to perform respiratory flow-phase triggered closed-loop diaphragmatic stimulation with resultant changes in respiratory flow and tidal volume.

Effects of spinal cord injury-induced changes in muscle activation on foot drag in a computational rat ankle model

Spinal cord injury (SCI) can lead to changes in muscle activation patterns and atrophy of affected muscles. Moderate levels of SCI are typically associated with foot drag during the swing phase of locomotion. Foot drag is often used to assess locomotor recovery, but the causes remain unclear. We hypothesized that foot drag results from inappropriate muscle coordination preventing flexion at the stance-to-swing transition. To test this hypothesis and to assess the relative contributions of neural and muscular changes on foot drag, we developed a two-dimensional, one degree of freedom ankle musculoskeletal model with gastrocnemius and tibialis anterior muscles. Anatomical data collected from sham-injured and incomplete SCI (iSCI) female Long-Evans rats as well as physiological data from the literature were used to implement an open-loop muscle dynamics model. Muscle insertion point motion was calculated with imposed ankle trajectories from kinematic analysis of treadmill walking in sham-injured and iSCI animals. Relative gastrocnemius deactivation and tibialis anterior activation onset times were varied within physiologically relevant ranges based on simplified locomotor electromyogram profiles. No-atrophy and moderate muscle atrophy as well as normal and injured muscle activation profiles were also simulated. Positive moments coinciding with the transition from stance to swing phase were defined as foot swing and negative moments as foot drag. Whereas decreases in activation delay caused by delayed gastrocnemius deactivation promote foot drag, all other changes associated with iSCI facilitate foot swing. Our results suggest that even small changes in the ability to precisely deactivate the gastrocnemius could result in foot drag after iSCI.

Adaptive control of ventilation using electrical stimulation in a biomechanical model

Cervical spinal cord injury (SCI) causes loss or impairment of control of respiratory muscles. Life-sustaining ventilation can be provided by mechanical ventilators (which have numerous side effects) or open-loop electrical stimulation respiratory pacing systems [1]. The use of adaptive control strategies in respiratory pacing systems can simplify initial setup procedures and allow the system to adjust stimulation values to account for changes due to muscle fatigue and/or respiratory demand. We have implemented a neural network based adaptive controller [2] with a biomechanical model of human ventilator dynamics and diaphragm stimulation in Simulink/SimMechanics/ Matlab. The adaptive controller uses sensor information to automatically determine a stimulation pattern that will produce a pre-specified desired lung volume trajectory. The controller uses a two-stage pattern generator/pattern shaper (PG/PS) structure which has successfully controlled leg movements in human subjects [3] and rats [4] using neuromuscular electrical stimulation. The biomechanical model incorporates a physiologically realistic Hill-type muscle model and a damped spring with non-linear compliance using published parameters for muscle geometry [5]. The parameters of the biomechanical model (muscle mass and lung damping) were varied +/-20% to simulate variation across a population to yield 9 sets of parameters. The quality of control and rate of adaptation achieved by the PG/PS controller were quantified by assessing the tracking error (difference between the actual and desired volume patterns) and the number of cycles needed to reach 5% error. Controller parameters were initialized to provide a nominal degree of ventilation during initial breaths. For each set of biomechanical parameters, the controller adapted stimulation values to achieve the same desired volume trajectory without any modification of initial controller values (Figure ​(Figure1)1) and to achieve less than 5% error in 1-10 (mean 5.4 ± 2.6) cycles. This adaptive strategy will be investigated further in simulation as well as hardware implementations for testing in animal models. Figure 1 A: Stimulation output for each biomechanical model tested for the first 10 breaths. B: Lung volume trajectories for each biomechanical model. The trial on the system with the weakest muscle and greatest load (damping) is shown in red. Note that all models ...

Effects of muscle strength and activation profile on foot drag in a simulated SCI rat

Introduction Spinal cord injury (SCI) can lead to decreases in overall motoneuron output, changes in muscle activation patterns, and atrophy of the affected muscles. Depending upon the severity of SCI, foot drag can occur during the stance-to-swing transition during locomotion. Our preliminary experimental data suggests that muscles co-activate following SCI, which may cause foot drag. To assess the relative roles of neural and muscular changes, a 3D neuro-musculoskeletal model of the rat hindlimb is being developed. Initially, a 2D, one degree-of-freedom ankle model with gastrocnemius (GAS) and tibialis anterior (TA) muscles was developed. Given that GAS can produce more ankle torque than TA, this simulation study tested the hypothesis that GAS motoneuronal deactivation profile dominates the ability of the foot to transition from stance to swing. Our results suggest that even small changes in the ability to precisely de-activate muscles could lead to kinematic deficits commonly observed following SCI.