Anatomical Registration of Implanted Sensors Improves Accuracy of Trunk Tilt Estimates with a Networked Neuroprosthesis.

For individuals with spinal cord injuries (SCIs) above the midthoracic level, a common complication is the partial or complete loss of trunk stability in the seated position. Functional neuromuscular stimulation (FNS) can restore seated posture and other motor functions after paralysis by applying small electrical currents to the peripheral motor nerves. In particular, the Networked Neuroprosthesis (NNP) is a fully implanted, modular FNS system that is also capable of capturing information from embedded accelerometers for measuring trunk tilt for feedback control of stimulation. The NNP modules containing the accelerometers are located in the body based on surgical constraints. As such, their exact orientations are generally unknown and cannot be easily assessed. In this study, a method for estimating trunk tilt that employed the Gram-Schmidt method to reorient acceleration signals to the anatomical axes of the body was developed and deployed in individuals with SCI using the implanted NNP system. An anatomically realistic model of a human trunk and five accelerometer sensors was developed to verify the accuracy of the reorientation algorithm. Correlation coefficients and root mean square errors (RMSEs) were calculated to compare target trunk tilt estimates and tilt estimates derived from simulated accelerometer signals under a variety of conditions. Simulated trunk tilt estimates with correlation coefficients above 0.92 and RMSEs below 5° were achieved. The algorithm was then applied to accelerometer signals from implanted sensors installed in three NNP recipients. Error analysis was performed by comparing the correlation coefficients and RMSEs derived from trunk tilt estimates calculated from implanted sensor signals to those calculated via motion capture data, which served as the gold standard. NNP-derived trunk tilt estimates exhibited correlation coefficients between 0.80 and 0.95 and RMSEs below 13° for both pitch and roll in most cases. These findings suggest that the algorithm is effective at estimating trunk tilt with the implanted sensors of the NNP system, which implies that the method may be appropriate for extracting feedback signals for control systems for seated stability with NNP technology for individuals who have reduced control of their trunk due to paralysis.

Robust Control of the Human Trunk Posture Using Functional Neuromuscular Stimulation: A Simulation Study.

The trunk movements of an individual paralyzed by spinal cord injury (SCI) can be restored by functional neuromuscular stimulation (FNS), which applies low-level current to the motor nerves to activate the paralyzed muscles to generate useful torques, to actuate the trunk. FNS can be modulated to vary the biotorques to drive the trunk to follow a user-defined reference motion and maintain it at a desired postural set-point. However, a stabilizing modulation policy (i.e., control law) is difficult to derive as the biomechanics of the spine and pelvis are complex and the neuromuscular dynamics are highly nonlinear, nonautonomous, and input redundant. Therefore, a control method that can stabilize it with FNS without knowing the accurate skeletal and neuromuscular dynamics is desired. To achieve this goal, we propose a control framework consisting of a robust control module that generates stabilizing torques while an artificial neural network-based mapping mechanism with an anatomy-based updating law ensures that the muscle-generated torques converge to the stabilizing values. For the robust control module, two sliding-mode robust controllers (i.e., a high compensation controller and an adaptive controller), were investigated. System stability of the proposed control method was rigorously analyzed based on the assumption that the skeletal dynamics can be approximated by Euler-Lagrange equations with bounded disturbances, which enables the generalization of the control framework. We present experiments in a simulation environment where an anatomically realistic three-dimensional musculoskeletal model of the human trunk moved in the anterior- posterior and medial-lateral directions while perturbations were applied. The satisfactory simulation results suggest the potential of this control technique for trunk tracking tasks in a typical clinical environment.

Feedback control of upright seating with functional neuromuscular stimulation during a reaching task after spinal cord injury: a feasibility study.

BACKGROUND: Restoring or improving seated stability after spinal cord injury (SCI) can improve the ability to perform activities of daily living by providing a dynamic, yet stable, base for upper extremity motion. Seated stability can be obtained with activation of the otherwise paralyzed trunk and hip musculature with neural stimulation, which has been shown to extend upper limb reach and improve seated posture. METHODS: We implemented a proportional, integral, derivative (PID) controller to maintain upright seated posture by simultaneously modulating both forward flexion and lateral bending with functional neuromuscular stimulation. The controller was tested with a functional reaching task meant to require trunk movements and impart internal perturbations through rapid changes in inertia due to acquiring, moving, and replacing objects with one upper extremity. Five subjects with SCI at various injury levels who had received implanted stimulators targeting their trunk and hip muscles participated in the study. Each subject was asked to move a weighted jar radially from a center home station to one of three target stations. The task was performed with the controller active, inactive, or with a constant low level of neural stimulation. Trunk pitch (flexion) and roll (lateral bending) angles were measured with motion capture and plotted against each other to generate elliptical movement profiles for each task and condition. Postural sway was quantified by calculating the ellipse area. Additionally, the mean effective reach (distance between the shoulder and wrist) and the time required to return to an upright posture was determined during reaching movements. RESULTS: Postural sway was reduced by the controller in two of the subjects, and mean effective reach was increased in three subjects and decreased for one. Analysis of the major direction of motion showed return to upright movements were quickened by 0.17 to 0.32 s. A 15 to 25% improvement over low/no stimulation was observed for four subjects. CONCLUSION: These results suggest that feedback control of neural stimulation is a viable way to maintain upright seated posture by facilitating trunk movements necessary to complete reaching tasks in individuals with SCI. Replication of these findings on a larger number of subjects would be necessary for generalization to the various segments of the SCI population.

Trunk Posture from Randomly Oriented Accelerometers.

Feedback control of functional neuromuscular stimulation has the potential to improve daily function for individuals with spinal cord injuries (SCIs) by enhancing seated stability. Our fully implanted networked neuroprosthesis (NNP) can provide real-time feedback signals for controlling the trunk through accelerometers embedded in modules distributed throughout the trunk. Typically, inertial sensors are aligned with the relevant body segment. However, NNP implanted modules are placed according to surgical constraints and their precise locations and orientations are generally unknown. We have developed a method for calibrating multiple randomly oriented accelerometers and fusing their signals into a measure of trunk orientation. Six accelerometers were externally attached in random orientations to the trunks of six individuals with SCI. Calibration with an optical motion capture system resulted in RMSE below 5° and correlation coefficients above 0.97. Calibration with a handheld goniometer resulted in RMSE of 7° and correlation coefficients above 0.93. Our method can obtain trunk orientation from a network of sensors without a priori knowledge of their relationships to the body anatomical axes. The results of this study will be invaluable in the design of feedback control systems for stabilizing the trunk of individuals with SCI in combination with the NNP implanted technology.

Estimating Center of Mass Kinematics During Perturbed Human Standing Using Accelerometers.

Estimating center of mass (COM) through sensor measurements is done to maintain walking and standing stability with exoskeletons. The authors present a method for estimating COM kinematics through an artificial neural network, which was trained by minimizing the mean squared error between COM displacements measured by a gold-standard motion capture system and recorded acceleration signals from body-mounted accelerometers. A total of 5 able-bodied participants were destabilized during standing through: (1) unexpected perturbations caused by 4 linear actuators pulling on the waist and (2) volitionally moving weighted jars on a shelf. Each movement type was averaged across all participants. The algorithm's performance was quantified by the root mean square error and coefficient of determination (R2) calculated from both the entire trial and during each perturbation type. Throughout the trials and movement types, the average coefficient of determination was 0.83, with 89% of the movements with R2 > .70, while the average root mean square error ranged between 7.3% and 22.0%, corresponding to 0.5- and 0.94-cm error in both the coronal and sagittal planes. COM can be estimated in real time for balance control of exoskeletons for individuals with a spinal cord injury, and the procedure can be generalized for other gait studies.

Automatic application of neural stimulation during wheelchair propulsion after SCI enhances recovery of upright sitting from destabilizing events.

BACKGROUND: The leading cause of injury for manual wheelchair users are tips and falls caused by unexpected destabilizing events encountered during everyday activities. The purpose of this study was to determine the feasibility of automatically restoring seated stability to manual wheelchair users with spinal cord injury (SCI) via a threshold-based system to activate the hip and trunk muscles with electrical stimulation during potentially destabilizing events. METHODS: We detected and classified potentially destabilizing sudden stops and turns with a wheelchair-mounted wireless inertial measurement unit (IMU), and then applied neural stimulation to activate the appropriate muscles to resist trunk movement and restore seated stability. After modeling and preliminary testing to determine the appropriate inertial signatures to discriminate between events and reliably trigger stimulation, the system was implemented and evaluated in real-time on manual wheelchair users with SCI. Three participants completed simulated collision events and four participants completed simulated rapid turns. Data were analyzed as a series of individual case studies with subjects acting as their own controls with and without the system active. RESULTS: The controller achieved 93% accuracy in detecting collisions and right turns, and 100% accuracy in left turn detection. Two of the three subjects who participated in collision testing with stimulation experienced significantly decreased maximum anterior-posterior trunk angles (p < 0.05). Similar results were obtained with implanted and surface stimulation systems. CONCLUSIONS: This study demonstrates the feasibility of a neural stimulation control system based on simple inertial measurements to improve trunk stability and overall safety of people with spinal cord injuries during manual wheelchair propulsion. Further studies are required to determine clinical utility in real world situations and generalizability to the broader SCI or other population of manual or powered wheelchair users. TRIAL REGISTRATION: ClinicalTrials.gov Identifier NCT01474148 . Registered 11/08/2011 retrospectively registered.

Reactive stepping with functional neuromuscular stimulation in response to forward-directed perturbations.

BACKGROUND: Implanted motor system neuroprostheses can be effective at increasing personal mobility of persons paralyzed by spinal cord injuries. However, currently available neural stimulation systems for standing employ patterns of constant activation and are unreactive to changing postural demands. METHODS: In this work, we developed a closed-loop controller for detecting forward-directed body disturbances and initiating a stabilizing step in a person with spinal cord injury. Forward-directed pulls at the waist were detected with three body-mounted triaxial accelerometers. A finite state machine was designed and tested to trigger a postural response and apply stimulation to appropriate muscles so as to produce a protective step when the simplified jerk signal exceeded predetermined thresholds. RESULTS: The controller effectively initiated steps for all perturbations with magnitude between 10 and 17.5 s body weight, and initiated a postural response with occasional steps at 5% body weight. For perturbations at 15 and 17.5% body weight, the dynamic responses of the subject exhibited very similar component time periods when compared with able-bodied subjects undergoing similar postural perturbations. Additionally, the reactive step occurred faster for stronger perturbations than for weaker ones (p < .005, unequal varience t-test.) CONCLUSIONS: This research marks progress towards a controller which can improve the safety and independence of persons with spinal cord injury using implanted neuroprostheses for standing.

A muscle-driven approach to restore stepping with an exoskeleton for individuals with paraplegia.

BACKGROUND: Functional neuromuscular stimulation, lower limb orthosis, powered lower limb exoskeleton, and hybrid neuroprosthesis (HNP) technologies can restore stepping in individuals with paraplegia due to spinal cord injury (SCI). However, a self-contained muscle-driven controllable exoskeleton approach based on an implanted neural stimulator to restore walking has not been previously demonstrated, which could potentially result in system use outside the laboratory and viable for long term use or clinical testing. In this work, we designed and evaluated an untethered muscle-driven controllable exoskeleton to restore stepping in three individuals with paralysis from SCI. METHODS: The self-contained HNP combined neural stimulation to activate the paralyzed muscles and generate joint torques for limb movements with a controllable lower limb exoskeleton to stabilize and support the user. An onboard controller processed exoskeleton sensor signals, determined appropriate exoskeletal constraints and stimulation commands for a finite state machine (FSM), and transmitted data over Bluetooth to an off-board computer for real-time monitoring and data recording. The FSM coordinated stimulation and exoskeletal constraints to enable functions, selected with a wireless finger switch user interface, for standing up, standing, stepping, or sitting down. In the stepping function, the FSM used a sensor-based gait event detector to determine transitions between gait phases of double stance, early swing, late swing, and weight acceptance. RESULTS: The HNP restored stepping in three individuals with motor complete paralysis due to SCI. The controller appropriately coordinated stimulation and exoskeletal constraints using the sensor-based FSM for subjects with different stimulation systems. The average range of motion at hip and knee joints during walking were 8.5°-20.8° and 14.0°-43.6°, respectively. Walking speeds varied from 0.03 to 0.06 m/s, and cadences from 10 to 20 steps/min. CONCLUSIONS: A self-contained muscle-driven exoskeleton was a feasible intervention to restore stepping in individuals with paraplegia due to SCI. The untethered hybrid system was capable of adjusting to different individuals' needs to appropriately coordinate exoskeletal constraints with muscle activation using a sensor-driven FSM for stepping. Further improvements for out-of-the-laboratory use should include implantation of plantar flexor muscles to improve walking speed and power assist as needed at the hips and knees to maintain walking as muscles fatigue.

A neuroprosthesis for control of seated balance after spinal cord injury.

BACKGROUND: A major desire of individuals with spinal cord injury (SCI) is the ability to maintain a stable trunk while in a seated position. Such stability is invaluable during many activities of daily living (ADL) such as regular work in the home and office environments, wheelchair propulsion and driving a vehicle. Functional neuromuscular stimulation (FNS) has the ability to restore function to paralyzed muscles by application of measured low-level currents to the nerves serving those muscles. METHODS: A feedback control system for maintaining seated balance under external perturbations was designed and tested in individuals with thoracic and cervical level spinal cord injuries. The control system relied on a signal related to the tilt of the trunk from the vertical position (which varied between 1.0 ≡ erect posture and 0.0 ≡ most forward flexed posture) derived from a sensor fixed to the sternum to activate the user's own hip and trunk extensor muscles via an implanted neuroprosthesis. A proportional-derivative controller modulated stimulation between trunk tilt values indicating deviation from the erect posture and maximum desired forward flexion. Tests were carried out with external perturbation forces set at 35%, 40% and 45% body-weight (BW) and maximal forward trunk tilt flexion thresholds set at 0.85, 0.75 and 0.70. RESULTS: Preliminary tests in a case series of five subjects show that the controller could maintain trunk stability in the sagittal plane for perturbations up to 45% of body weight and for flexion thresholds as low as 0.7. The mean settling time varied across subjects from 0.5(±0.4) and 2.0 (±1.1) seconds. Mean response time of the feedback control system varied from 393(±38) ms and 536(±84) ms across the cohort. CONCLUSIONS: The results show the high potential for robust control of seated balance against nominal perturbations in individuals with spinal cord injury and indicates that trunk control with FNS is a promising intervention for individuals with SCI.

Intrinsic and Extrinsic Contributions to Seated Balance in the Sagittal and Coronal Planes: Implications for Trunk Control After Spinal Cord Injury.

The contributions of intrinsic (passive) and extrinsic (active) properties of the human trunk, in terms of the simultaneous actions about the hip and spinal joints, to the control of sagittal and coronal seated balance were examined. Able-bodied (ABD) and spinal-cord-injured (SCI) volunteers sat on a moving platform which underwent small amplitude perturbations in the anterior-posterior (AP) and medial-lateral (ML) directions while changes to trunk orientation were measured. A linear parametric model that related platform movement to trunk angle was fit to the experimental data by identifying model parameters in the time domain. The results showed that spinal cord injury leads to a systematic reduction in the extrinsic characteristics, while most of the intrinsic characteristics were rarely affected. In both SCI and ABD individuals, passive characteristics alone were not enough to maintain seated balance. Passive stiffness in the ML direction was almost 3 times that in the AP direction, making more extrinsic mechanisms necessary for balance in the latter direction. Proportional and derivative terms of the extrinsic model made the largest contribution to the overall output from the active system, implying that a simple proportional plus derivative (PD) controller structure will suffice for restoring seated balance after spinal cord injury.

Center of Mass Estimation for Impaired Gait Assessment Using Inertial Measurement Units.

Injury or disease often compromise walking dynamics and negatively impact quality of life and independence. Assessing methods to restore or improve pathological gait can be expedited by examining a global parameter that reflects overall musculoskeletal control. Center of mass (CoM) kinematics follow well-defined trajectories during unimpaired gait, and change predictably with various gait pathologies. We propose a method to estimate CoM trajectories from inertial measurement units (IMUs) using a bidirectional Long Short-Term Memory neural network to evaluate rehabilitation interventions and outcomes. Five non-disabled volunteers participated in a single session of various dynamic walking trials with IMUs mounted on various body segments. A neural network trained with data from four of the five volunteers through a leave-one-subject out cross validation estimated the CoM with average root mean square errors (RMSEs) of 1.44cm, 1.15cm, and 0.40cm in the mediolateral (ML), anteroposterior (AP), and inferior/superior (IS) directions respectively. The impact of number and location of IMUs on network prediction accuracy was determined via principal component analysis. Comparing across all configurations, three to five IMUs located on the legs and medial trunk were the most promising reduced sensor sets for achieving CoM estimates suitable for outcome assessment. Lastly, the networks were tested on data from an individual with hemiparesis with the greatest error increase in the ML direction, which could stem from asymmetric gait. These results provide a framework for assessing gait deviations after disease or injury and evaluating rehabilitation interventions intended to normalize gait pathologies.

Effects of stimulating hip and trunk muscles on seated stability, posture, and reach after spinal cord injury.

OBJECTIVE: To determine the stimulated strength of the paralyzed gluteal and paraspinal muscles and their effects on the seated function of individuals with paralysis. DESIGN: Case series with subjects acting as their own concurrent controls. SETTING: Hospital-based clinical biomechanics laboratory. PARTICIPANTS: Users (N=8) of implanted neuroprostheses for lower extremity function with low-cervical or thoracic level injuries. INTERVENTIONS: Dynamometry and digital motion capture both with and without stimulation to the hip and trunk muscles. MAIN OUTCOME MEASURES: Isometric trunk extension moment at 0°, 15°, and 30° of flexion; seated stability in terms of simulated isokinetic rowing; pelvic tilt, shoulder height, loaded and unloaded bimanual reaching to different heights, and subjective ratings of difficulty during unsupported sitting. RESULTS: Stimulation produced significant increases in mean trunk extension moment (9.2±9.5Nm, P<.001) and rowing force (27.4±23.1N, P<.012) over baseline volitional values. Similarly, stimulation induced positive changes in average pelvic tilt (16.7±15.7°) and shoulder height (2.2±2.5cm) during quiet sitting and bimanual reaching, and increased mean reach distance (5.5±6.6cm) over all subjects, target heights, and loading conditions. Subjects consistently rated tasks with stimulation easier than voluntary effort alone. CONCLUSIONS: In spite of considerable intersubject variability, stabilizing the paralyzed trunk with electrical stimulation can positively impact seated posture, extend forward reach, and allow exertion of larger forces on objects in the environment.

Effects of intramuscular trunk stimulation on manual wheelchair propulsion mechanics in 6 subjects with spinal cord injury.

OBJECTIVE: To quantify the effects of stabilizing the paralyzed trunk and pelvis with electrical stimulation on manual wheelchair propulsion. DESIGN: Single-subject design case series with subjects acting as their own concurrent controls. SETTING: Hospital-based clinical biomechanics laboratory. PARTICIPANTS: Individuals (N=6; 4 men, 2 women; mean age ± SD, 46 ± 10.8y) who were long-time users (6.1 ± 3.9y) of implanted neuroprostheses for lower extremity function and had chronic (8.6 ± 2.8y) midcervical- or thoracic-level injuries (C6-T10). INTERVENTIONS: Continuous low-level stimulation to the hip (gluteus maximus, posterior adductor, or hamstrings) and trunk extensor (lumbar erector spinae and/or quadratus lumborum) muscles with implanted intramuscular electrodes. MAIN OUTCOME MEASURES: Pushrim kinetics (peak resultant force, fraction effective force), kinematics (cadence, stroke length, maximum forward lean), and peak shoulder moment at preferred speed over 10-m level surface; speed, pushrim kinetics, and subjective ratings of effort for level 100-m sprints and up a 30.5-m ramp of approximately 5% grade. RESULTS: Three of 5 subjects demonstrated reduced peak resultant pushrim forces (P≤.014) and improved efficiency (P≤.048) with stimulation during self-paced level propulsion. Peak sagittal shoulder moment remained unchanged in 3 subjects and increased in 2 others (P<.001). Maximal forward trunk lean also increased by 19% to 26% (P<.001) with stimulation in these 3 subjects. Stroke lengths were unchanged by stimulation in all subjects, and 2 showed extremely small (5%) but statistically significant increases in cadence (P≤.021). Performance measures for sprints and inclines were generally unchanged with stimulation; however, subjects consistently rated propulsion with stimulation to be easier for both surfaces. CONCLUSIONS: Stabilizing the pelvis and trunk with low levels of continuous electrical stimulation to the lumbar trunk and hip extensors can positively impact the mechanics of manual wheelchair propulsion and reduce both perceived and physical measures of effort.

Stance controlled knee flexion improves stimulation driven walking after spinal cord injury.

BACKGROUND: Functional neuromuscular stimulation (FNS) restores walking function after paralysis from spinal cord injury via electrical activation of muscles in a coordinated fashion. Combining FNS with a controllable orthosis to create a hybrid neuroprosthesis (HNP) has the potential to extend walking distance and time by mechanically locking the knee joint during stance to allow knee extensor muscle to rest with stimulation turned off. Recent efforts have focused on creating advanced HNPs which couple joint motion (e.g., hip and knee or knee and ankle) to improve joint coordination during swing phase while maintaining a stiff-leg during stance phase. METHODS: The goal of this study was to investigate the effects of incorporating stance controlled knee flexion during loading response and pre-swing phases on restored gait. Knee control in the HNP was achieved by a specially designed variable impedance knee mechanism (VIKM). One subject with a T7 level spinal cord injury was enrolled and served as his own control in examining two techniques to restore level over-ground walking: FNS-only (which retained a stiff knee during stance) and VIKM-HNP (which allowed controlled knee motion during stance). The stimulation pattern driving the walking motion remained the same for both techniques; the only difference was that knee extensor stimulation was constant during stance with FNS-only and modulated together with the VIKM to control knee motion during stance with VIKM-HNP. RESULTS: Stance phase knee angle was more natural during VIKM-HNP gait while knee hyperextension persisted during stiff-legged FNS-only walking. During loading response phase, vertical ground reaction force was less impulsive and instantaneous gait speed was increased with VIKM-HNP, suggesting that knee flexion assisted in weight transfer to the leading limb. Enhanced knee flexion during pre-swing phase also aided flexion during swing, especially when response to stimulation was compromised. CONCLUSIONS: These results show the potential advantages of incorporating stance controlled knee flexion into a hybrid neuroprosthesis for walking. The addition of such control to FNS driven walking could also enable non-level walking tasks such as uneven terrain, slope navigation and stair descent where controlled knee flexion during weight bearing is critical.

Stabilizing leaning postures with feedback controlled functional neuromuscular stimulation after trunk paralysis.

Spinal cord injury (SCI) can cause paralysis of trunk and hip musculature that negatively impacts seated balance and ability to lean away from an upright posture and interact fully with the environment. Constant levels of electrical stimulation of peripheral nerves can activate typically paralyzed muscles and aid in maintaining a single upright seated posture. However, in the absence of a feedback controller, such seated postures and leaning motions are inherently unstable and unable to respond to perturbations. Three individuals with motor complete SCI who had previously received a neuroprosthesis capable of activating the hip and trunk musculature volunteered for this study. Subject-specific muscle synergies were identified through system identification of the lumbar moments produced via neural stimulation. Synergy-based calculations determined the real-time stimulation parameters required to assume leaning postures. When combined with a proportional, integral, derivative (PID) feedback controller and an accelerometer to infer trunk orientation, all individuals were able to assume non-erect postures of 30-40° flexion and 15° lateral bending. Leaning postures increased forward reaching capabilities by 10.2, 46.7, and 16 cm respectively for each subject when compared with no stimulation. Additionally, the leaning controllers were able to resist perturbations of up to 90 N, and all subjects perceived the leaning postures as moderately to very stable. Implementation of leaning controllers for neuroprostheses have the potential of expanding workspaces, increasing independence, and facilitating activities of daily living for individuals with paralysis.

Comparing joint kinematics and center of mass acceleration as feedback for control of standing balance by functional neuromuscular stimulation.

BACKGROUND: The purpose of this study was to determine the comparative effectiveness of feedback control systems for maintaining standing balance based on joint kinematics or total body center of mass (COM) acceleration, and assess their clinical practicality for standing neuroprostheses after spinal cord injury (SCI). METHODS: In simulation, controller performance was measured according to the upper extremity effort required to stabilize a three-dimensional model of bipedal standing against a variety of postural disturbances. Three cases were investigated: proportional-derivative control based on joint kinematics alone, COM acceleration feedback alone, and combined joint kinematics and COM acceleration feedback. Additionally, pilot data was collected during external perturbations of an individual with SCI standing with functional neuromuscular stimulation (FNS), and the resulting joint kinematics and COM acceleration data was analyzed. RESULTS: Compared to the baseline case of maximal constant muscle excitations, the three control systems reduced the mean upper extremity loading by 51%, 43% and 56%, respectively against external force-pulse perturbations. Controller robustness was defined as the degradation in performance with increasing levels of input errors expected with clinical deployment of sensor-based feedback. At error levels typical for body-mounted inertial sensors, performance degradation due to sensor noise and placement were negligible. However, at typical tracking error levels, performance could degrade as much as 86% for joint kinematics feedback and 35% for COM acceleration feedback. Pilot data indicated that COM acceleration could be estimated with a few well-placed sensors and efficiently captures information related to movement synergies observed during perturbed bipedal standing following SCI. CONCLUSIONS: Overall, COM acceleration feedback may be a more feasible solution for control of standing with FNS given its superior robustness and small number of inputs required.

Trunk acceleration for neuroprosthetic control of standing: a pilot study.

This pilot study investigated the potential of using trunk acceleration feedback control of center of pressure (COP) against postural disturbances with a standing neuroprosthesis following paralysis. Artificial neural networks (ANNs) were trained to use three-dimensional trunk acceleration as input to predict changes in COP for able-bodied subjects undergoing perturbations during bipedal stance. Correlation coefficients between ANN predictions and actual COP ranged from 0.67 to 0.77. An ANN trained across all subject-normalized data was used to drive feedback control of ankle muscle excitation levels for a computer model representing a standing neuroprosthesis user. Feedback control reduced average upper-body loading during perturbation onset and recovery by 42% and peak loading by 29% compared with optimal, constant excitation.

Facilitation of dependent transfers with functional neuromuscular stimulation: a computer simulation study.

A two-part simulation process was developed to investigate the facilitation of vertical patient lifts with functional neuromuscular stimulation (FNS) in individuals with spinal cord injury (SCI). First, external lifting forces representing caregiver assistance were applied to a 3D musculoskeletal model representing the patient and optimized to enforce a specific lifting trajectory during a forward dynamic simulation. The process was repeated with and without the activation of the knee, hip, and trunk extensor muscles of the patient model to represent contractions of the paralyzed muscles generated via FNS. Secondly, the spinal compression experienced by a caregiver at the L5/S1 joint while generating these external lifting forces was estimated using a second musculoskeletal model representing the caregiver. Simulation without muscle activation predicted spinal compression in the caregiver model approximately 1.3 × the National Institute for Occupational Safety and Health (NIOSH) recommended "Action Limit." Comparatively, simulations with two unique patterns of muscle activation both predicted caregiver spinal compressions below NIOSH recommendations. These simulation results support the hypothesis that FNS activation of a patient's otherwise paralyzed muscles would lower the force output required of a caregiver during a dependent transfer, thus lowering the spinal compression and risk of injury experienced by a caregiver.

Neural Networks Trained via Reinforcement Learning Stabilize Walking of a Three-Dimensional Biped Model With Exoskeleton Applications.

Our group is developing a cyber-physical walking system (CPWS) for people paralyzed by spinal cord injuries (SCI). The current CPWS consists of a functional neuromuscular stimulation (FNS) system and a powered lower-limb exoskeleton for walking with leg movements in the sagittal plane. We are developing neural control systems that learn to assist the user of this CPWS to walk with stability. In a previous publication (Liu et al., Biomimetics, 2019, 4, 28), we showed a neural controller that stabilized a simulated biped in the sagittal plane. We are considering adding degrees of freedom to the CPWS to allow more natural walking movements and improved stability. Thus, in this paper, we present a new neural network enhanced control system that stabilizes a three-dimensional simulated biped model of a human wearing an exoskeleton. Results show that it stabilizes human/exoskeleton models and is robust to impact disturbances. The simulated biped walks at a steady pace in a range of typical human ambulatory speeds from 0.7 to 1.3 m/s, follows waypoints at a precision of 0.3 m, remains stable, and continues walking forward despite impact disturbances and adapts its speed to compensate for persistent external disturbances. Furthermore, the neural network controller stabilizes human models of different statures from 1.4 to 2.2 m tall without any changes to the control parameters. Please see videos at the following link: 3D biped walking control.

Characterization of the Force Production Capabilities of Paralyzed Trunk Muscles Activated With Functional Neuromuscular Stimulation in Individuals With Spinal Cord Injury.

Paralysis of the trunk results in seated instability leading to difficulties performing activities of daily living. Functional neuromuscular stimulation (FNS) combined with control systems have the potential to restore some dynamic functions of the trunk. However, design of multi-joint, multi-muscle control systems requires characterization of the stimulation-driven muscles responsible for movement. OBJECTIVE: This study characterizes the input-output properties of paralyzed trunk muscles activated by FNS, and explores co-activation of muscles. METHODS: Four participants with various spinal cord injuries (C7 AIS-B, T4 AIS-B, T5 AIS-A, C5 AIS-C) were constrained so lumbar forces were transmitted to a load cell while an implanted neuroprosthesis activated otherwise paralyzed hip and paraspinal muscles. Isometric force recruitment curves in the nominal seated position were generated by inputting the level of stimulation (pulse width modulation) while measuring the resulting muscle force. Two participants returned for a second experiment where muscles were co-activated to determine if their actions combined linearly. RESULTS: Recruitment curves of most trunk and hip muscles fit sigmoid shaped curves with a regression coefficient above 0.75, and co-activation of the muscles combined linearly across the hip and lumbar joint. Subject specific perturbation plots showed one subject is capable of resisting up to a 300N perturbation anteriorly and 125N laterally; with some subjects falling considerably below these values. CONCLUSION: Development of a trunk stability control system can use sigmoid recruitment dynamics and assume muscle forces combine linearly. SIGNIFICANCE: This study informs future designs of multi-muscle, and multi-dimensional FNS systems to maintain seated posture and stability.