Improving signal reliability for on-line joint angle estimation from nerve cuff recordings of muscle afferents

Closed-loop functional electrical stimulation (FES) applications depend on sensory feedback, thus, it is important to continuously investigate new methods to obtain reliable feedback signals. The objective of the present paper was to examine the feasibility of using an artificial neural network (ANN) to predict joint angle from whole nerve cuff recordings of muscle afferent activity within a physiological range of motion. Furthermore, we estimated how small changes in joint angle that can be detected from the nerve cuff recordings. Neural networks were tested with data obtained from ten acute rabbit experiments in simulated, on-line experiments. The electroneurograms (ENG) of the tibial and peroneal nerves were recorded during passive ankle joint rotation. To decrease the joint angle prediction error with new rabbit data, we attempted to pretune the nerve signals and re-trained the ANNs with the pretuned data. With these procedures we were able to compensate for interrabbit variability. On average the mean prediction errors were less than 2.0/spl deg/ (a total excursion of 20/spl deg/) and we were able to predict joint angles from muscle afferent activity with accuracy close to the best-estimated angular resolution. The angular resolution was found to depend on the initial joint angle and the actual step size taken and we found that there was a low probability of detecting joint angle changes less than 1.5/spl deg/. We thus suggest that muscle afferent activity is applicable as feedback in real-time closed-loop control, when the motion speed is restricted and when the movement is limited to a portion of the joint's physiological range.     

Nerve cuff recordings of muscle afferent activity from tibial and peroneal nerves in rabbit during passive ankle motion.

Activity from muscle afferents regarding ankle kinesthesia was recorded using cuff electrodes in a rabbit preparation in which tactile input from the foot was eliminated. The purpose was to determine if such activity can provide information useful in controlling functional electrical stimulation (FES) systems that restore mobility in spinal injured man. The rabbit's ankle was passively flexed and extended while the activity of the tibial and peroneal nerves was recorded. Responses to trapezoidal stimulus profiles were investigated for excursions from 10 degrees to 60 degrees using velocities from 5 degrees/s to 30 degrees/s and different initial ankle positions. The recorded signals mainly reflect activity from primary and secondary muscle afferents. Dorsiflexion stretched the ankle extensors and produced velocity dependent activity in the tibial nerve, and this diminished to a tonic level during the stimulus plateau. The peroneal nerve was silent during dorsiflexion, but was activated by stretch of the peroneal muscles during ankle extension. The responses of the two nerves behaved in a reciprocal manner, but exhibited considerable hysteresis, since motion that relaxed the stretch to the driving muscle produced an immediate cessation of the prior stretch induced activity. A noted difference between the tibial and peroneal nerve responses is that the range of joint position change that activated the flexor afferents was greater then for the extensor afferents. Ankle rotation at higher velocities increased the dynamic stretch evoked responses during the stimulus ramp but showed no effect on the tonic activity during the stimulus plateau. Prestretching the muscles by altering the initial position increased the response to the ramp movement, however, for the peroneal nerve, when the prestretch brought the flexor muscles near to their maximal lengths, the response to additional stretch provided by the ramp movement was diminished. The results indicate that the whole nerve recorded muscle afferent activity may be useful for control of FES assisted standing, because it can indicate the direction of rotation of the passively moved ankle joint, along with coarse information regarding the rate of movement and static joint position.     

Interpretation of Muscle Spindle Afferent Nerve Response to Passive Muscle Stretch Recorded With Thin-Film Longitudinal Intrafascicular Electrodes

In this study, we explored the feasibility of estimating muscle length in passive conditions by interpreting nerve responses from muscle spindle afferents recorded with thin-film longitudinal intrafascicular electrodes. Afferent muscle spindle response to passive stretch was recorded in ten acute rabbit experiments. A newly proposed first-order model of muscle spindle response to passive sinusoidal muscle stretch manages to capture the relationship between afferent neural firing rate and muscle length. We demonstrate that the model can be used to track random motion trajectories with bandwidth from 0.1 to 1 Hz over a range of 4 mm with a muscle length estimation error of 0.3 mm (1.4$^circ$ of joint angle). When estimation is performed using four-channel ENG there is a 50% reduction in estimate variation, compared to using single-channel recordings.      

Modeling study of peripheral nerve recording selectivity.

Recording of sensory information from afferent fibers can be used as feedback for the closed-loop control of neural prostheses. Clinical applications suggest that recording selectively from various nerve fascicles is important. Current nerve cuff electrodes are generally circular in shape and use a tripolar recording configuration. Preliminary experiments suggest that slowly changing the shape of the nerve to a flatter cross section can improve its selectivity. The objective of this work is to determine the effects of nerve reshaping and other cuff design parameters on the fascicular recording selectivity of a nerve cuff. A finite-element computer model of a multifasciculated nerve with different cuff electrodes was implemented to simulate the recordings. The model included the inhomogeneous and anisotropic properties of peripheral nerves. The recording selectivity was quantified with the use of a Selectivity Index. The results from the model provided information regarding the effect of using monopolar versus tripolar recording configurations, the length of the tripoles in tripolar recordings, the number of contacts that maximize the selectivity index, and the cuff length. Nerve reshaping was found to cause important recording selectivity improvements (106% average). These results provide specific criteria for the design of selectively recording nerve cuff electrodes.     

Comparison of three different models to represent the wrist duringwheelchair propulsion

Due to the high incidence of secondary wrist injury among manual wheelchair users, recent emphasis has been placed on the investigation of wheelchair propulsion biomechanics. Accurate representation of wrist activity during wheelchair propulsion may help to elucidate the mechanisms contributing to the development of wrist injuries. Unfortunately, no consensual wrist biomechanical model has been established. In order to determine if different methodologies obtain similar results, this investigation created and compared three different wrist models: 1) a fixed joint center placed between the styloids (midstyloid joint center); 2) a joint center with 2° of freedom computed from de Leva's joint center data; and 3) a floating joint center. Results indicate that wrist flexion and extension angles are highly consistent between models, however, radial and ulnar deviation angles vary considerably. Mean maximum right flexion angles were found to be 3.5°, 2.2°, and 5.0° for the midstyloid, de Leva, and floating joint center models, respectively. Extension angles were 22.3°, 23.6°, and 23.6°, respectively. Mean maximum right radial deviation angles for the midstyloid, de Leva, and floating joint center models were 26.0°, 26.9°, and 45.1°, respectively, and ulnar deviation angles were found to be 30.5°, 38.8°, and 10.2°, respectively. This information is useful when comparing kinematic studies and further supports the need for consensual methodology     

Restoration of use of paralyzed limb muscles using sensory nerve signals for state control of FES-assisted walking.

A real-time functional electrical stimulation (FES) state controller was designed that utilized sensory nerve cuff signals from the cat forelimb to control the timing of stimulation of the Palmaris Longus (PalL) muscle during walking on the treadmill. Sensory nerve signals from the median and superficial radial nerves provided accurate, reliable feedback related to foot contact and lift-off which, when analyzed with single threshold Schmitt triggers, produced valuable state information about the step cycle. The study involved three experiments: prediction of the timing of muscle activity in an open-loop configuration with no stimulation, prediction of the timing of muscle activity in a closed-loop configuration that included stimulation of the muscle over natural PaIL electromyogram (EMG), and temporary paralysis of selected forelimb muscles coupled with the use of the state controller to stimulate the PalL in order to return partial support function to the anesthetized limb. The FES state controller was tested in a variety of walking conditions, including different treadmill speeds and slopes. The results obtained in these experiments demonstrate that nerve cuff signals can provide a useful source of feedback to FES systems for control of limb function.     

Reflex regulation of antagonist muscles for control of joint equilibrium position.

A systems model of spinal neuro-musculo-skeletal system (alpha - gamma model) is developed to investigate the plausible roles of spinal proprioceptive feedback in movement control. The model is composed of a joint, a pair of antagonist muscles, length and velocity feedback from muscle spindle, as well as spinal stretch reflex, reciprocal inhibition and recurrent inhibition of Renshaw cells. A descending command modulates the background activation of alpha motoneuron pools in combination with these reflex activities. A static gamma command controls the fusimotor contraction of the spindle. Simulation results reveal that the equilibrium joint angle is linearly correlated to the level of static gamma fusimotor activity of the spindle for a wide range of external loading conditions and reflex gains, suggesting that these spinal reflexes may contribute to regulate the equilibrium position of the joint. Sensitivity analysis further shows that reflex gains and other central commands alter the quasi-linear relation in regular fashions. The reciprocal inhibition gain changes the slope of the linear theta(eq) - gamma curve; and the descending alpha excitation, the stretch reflex gain, and the external load all shift the theta(eq) - gamma curve in parallel. These results imply that reflex gains and descending alpha commands may be coordinated to maintain a unique theta(eq) - gamma curve while providing the flexibility to counteract external loads, to execute a movement, or to regulate additional muscle variables. Dynamic simulation suggests that control of a class of movements can be achieved with a triphasic, alpha pulse and a continuous gamma signal. The model study supports the notion of a dual strategy for controlling trajectories via a feedforward alpha command and for regulating the final equilibrium positions via a feedback gamma command.     

The development of a potential optimized stimulation intensity envelope for drop foot applications

An optimized stimulation intensity envelope for use in hemiplegic drop foot applications has been developed. The traditional trapezoidal stimulation intensity approach has been examined and found to be inconsistent with the muscle activity patterns observed in healthy gait and therefore unsuitable. Experimental functional electrical stimulation (FES)-elicited tibialis anterior (TA) electromyography (EMG) data was taken over the ankle range of interest (occurring during active dorsiflexion and loading response) while also taking into account the type of TA muscle contraction occurring (concentric, eccentric, and isometric) and the speed of hemiplegic ankle joint rotation. Using the processed data, a model of normalized EMG versus pulsewidth was developed. Implementation of this model showed the unsuitability of the trapezoidal approach in the reproducing of a natural EMG profile. An optimized stimulation intensity profile is proposed which is expected to accurately reproduce the natural TA EMG profile during gait.     

Comparison of joint torque evoked with monopolar and tripolar-cuff electrodes

Using a self-sizing spiral-cuff electrode placed on the sciatic nerve of the cat, the joint torque evoked with stimulation applied to contacts in a monopolar configuration was judged to be the same as the torque evoked by stimulation applied to contacts in a tripolar configuration. Experiments were carried out in six acute cat preparations. In each experiment, a 12-contact electrode was placed on the sciatic nerve and used to effect both the monopolar and tripolar electrode configurations. The ankle torque produced by electrically evoked isometric muscle contraction was measured in three dimensions: plantar flexion, internal rotation, and inversion. Based on the recorded ankle torque, qualitative and quantitative comparisons were performed to determine if any significant difference existed in the pattern or order in which motor nerve fibers were recruited. No significant difference was found at a 98% confidence interval in either the recruitment properties or the repeatability of the monopolar and tripolar configurations. Further, isolated activation of single fascicles within the sciatic nerve was observed. Once nerve fibers in a fascicle were activated, recruitment of that fascicle was modulated over the full range before "spill-over" excitation occurred in neighboring fascicles. These results indicate that a four contact, monopolar nerve-cuff electrode is a viable substitute for a 12 contact, tripolar nerve-cuff electrode. The results of this study are also consistent with the hypothesis that multicontact self-sizing spiral-cuff electrodes can be used in motor prostheses to provide selective control of many muscles. These findings should also apply to other neuroprostheses employing-cuff electrodes on nerve trunks.     

Direct neural sensory feedback and control of a prosthetic arm.

Evidence indicates that user acceptance of modern artificial limbs by amputees would be significantly enhanced by a system that provides appropriate, graded, distally referred sensations of touch and joint movement, and that the functionality of limb prostheses would be improved by a more natural control mechanism. We have recently demonstrated that it is possible to implant electrodes within individual fascicles of peripheral nerve stumps in amputees, that stimulation through these electrodes can produce graded, discrete sensations of touch or movement referred to the amputee's phantom hand, and that recordings of motor neuron activity associated with attempted movements of the phantom limb through these electrodes can be used as graded control signals. We report here that this approach allows amputees to both judge and set grip force and joint position in an artificial arm, in the absence of visual input, thus providing a substrate for better integration of the artificial limb into the amputee's body image. We believe this to be the first demonstration of direct neural feedback from and direct neural control of an artificial arm in amputees.     

Noninvasive measurement of torque development in the rat foot: measurement setup and results from stimulation of the sciatic nerve with polyimide-based cuff electrodes

In neural rehabilitation, selective activation of muscles after electrical stimulation is mandatory for control of paralyzed limbs. For an evaluation of electrode selectivity, a setup to noninvasively measure the force development after electrical stimulation in the rat foot was developed. The setup was designed in accordance to the anatomical features of the rat model to test the isometric torque development at given ankle positions in an intact leg. In this paper, the setup design and development is presented and discussed. In a first study, the selectivity of small nerve cuffs with 12 electrodes implanted around the rat sciatic nerve was investigated. Special attention was drawn to the performance of the torque measurement setup in comparison to electrophysiological data obtained from compound muscle action potential recordings. Using one cuff around the nerve, electrical stimulation on different electrode tripoles led to plantarflexion and dorsiflexion of the foot without an a priori alignment of the cuff.     

Variable patterned pudendal nerve stimuli improves reflex bladder activation.

We evaluated variable patterns of pudendal nerve (PN) stimuli for reflex bladder excitation. Reflex activation of the bladder has been demonstrated previously with 20-33 Hz continuous stimulation of PN afferents. Neuronal circuits accessed by afferent mediated pathways may respond better to physiological patterned stimuli than continuous stimulation. Unilateral PN nerve cuffs were placed in neurologically intact male cats. PN stimulation (0.5-100 Hz) was performed under isovolumetric conditions at bladder volumes up to the occurrence of distension evoked reflex contractions. Stimulus evoked reflex bladder contractions were elicited in eight cats. Across all experiments, bursting of 2-10 pulses at 100-200 Hz repeated at continuous stimulation frequencies evoked significantly larger bladder responses than continuous (single pulse) stimulation (52.0+/-44.5%). Bladder excitation was also effective at 1 Hz continuous stimuli, which is lower than typically reported. Variable patterned pulse bursting resulted in greater evoked reflex bladder pressures and increased the potential stimulation parameter space for effective bladder excitation. Improved bladder excitation should increase the efficacy of neuroprostheses for bladder control.     

Kinematic modeling for the assessment of wheelchair user'sstability

A computer kinematic model was developed to simulate the lateral and transverse stabilities of wheelchair users in order to compare the effect of different backrests. This model is composed of ellipsoids and parallelepipeds representing the main components of the human body, the seating devices and the wheelchair. A fifteen-segment three-dimensional (3-D) model linked by spherical and revolute joints was created using the ADAMS software (Mechanical Dynamics, Inc.). Torsional springs and dampers are used at the joints to represent four sets of articulation stiffness. Seating devices are represented with 45 rectangular surface patches. The interface between human body and seating devices is modeled by contact elements, which included the specification of stiffness, damping, and deformation of cushions and buttocks. Simulations of a user and his wheelchair moving at 1.4 m/s on a tilted pathway were performed. Different indexes [trunk lateral tilt (TLT) and trunk transverse rotation (TTR)] were measured and compared to those of a similar experimental study on four subjects. The effect of joint stiffness was quantified and a sensitivity study showed the importance of the hip, neck, lumbar, and thoracic joint stiffness on model response (between 16% and 68%). Two backrests (standard and highly contoured) were tested with the kinematic model and their stability compared. Overall, the coherence between the simulations and the experiments shows that this approach is appropriate to compare various seating devices (maximal difference of 1.3° between the simulated and experimental curves for the intermediate joint stiffness sets). The smallest rotations of the highly contoured backrest (6.3° versus 8.9° for TLT and 3.9° versus 6.7° for TTR) suggest that the contouring of the mid torso is more efficient than the lower torso to provide stability to the wheelchair user. This model is an adequate tool to test and improve the design of seating aids     

In search of optimality in sensorimotor control of standing posture

Summary form only given. If there are consistent criteria that the body utilizes to make choices, it may be possible to study human movement as an optimization problem: mathematical modeling of the criteria used to resolve redundancies. This problem can be broken into three components: characterization of the system being controlled, determination of optimization criteria for the control, and processing of sensory information that is used for feedback. It is shown that by examining the mapping between muscle activations and joint angular accelerations, it is possible to characterize the bounds on movement placed by muscle strength, musculoskeletal geometry, and the equations of motion for the body. The effect of these biomechanical factors is to constrain the set of joint angular accelerations that can be achieved given any combination of muscle activations. There are also additional constraints on body configurations and accelerations based on the desire to avoid falling over or lifting of the toes or heels off the ground. These effects are described by the set of all feasible joint angular accelerations     

Neurofuzzy adaptive controlling of selective stimulation for FES: a case study.

A controller was designed for the selective stimulation of the sciatic nerve with a multiple contact cuff electrode to generate a desired torque in the ankle joint of cat. The design integrates three approaches, artificial neural network (ANN) modeling, fuzzy logical adaptation, and geometrical mapping. The geometrical mapping refers to the vector transformation from the joint coordinates to the virtual muscle coordinates which have been conceptually developed to represent the major recruitment features of contact-based functional units in the physical plant. This method reduces the complexity of generating a data set for training the neural network in the feedforward path and implementing the on-line learning algorithm embedded in the feedback loop. The controller was evaluated by computer simulation with the experimental data obtained from the torque generation in five acute cats. The results show that the ANN-based feedforward is capable of predicting 65% of a given desired isometric torque, and the fuzzy logical machine is able to provide suitable gains for feedback modulation to reduce the error from 35 to 8.5% and produce a robust control.     

Cuff electrodes for long-term recording of natural sensoryinformation

Cuff electrodes for recording of the electro-neurogram from peripheral nerves were introduced by Hoffer [1974] and Stein, et al. [1975]. The cuffs were used to obtain higher signal amplitudes than previously possible, at least in chronic recordings, and to decrease the pick-up of noise, especially from muscles. Cuff electrodes are relatively stable in long-term recordings, but the stability has never been quantified in terms of input-output relationships; i.e., in terms of responses to repeatable stimuli over time. Moreover. The relationship between nerve damage and electrophysiological parameters has never been assessed. In this article, after reviewing the development of cuff electrodes and their applications, we present a long-term study of tactile peripheral nerve signals, electrically activated nerve signals, and impedance measurements. We show how the recordings vary over a 16-month period after implantation of nerve cuff electrodes in rabbits, and how nerve damage is reflected in the recorded signals     

Knee elasticity influenced by joint angle and perturbation intensity.

The responses of the human knee joint system to small rotational displacements were studied in the horizontal plane to eliminate the effect of gravity. Band-limited noise was used to produce up to +/-3 degrees angular perturbations to the knee joint. Under these restricted conditions, the knee joint system could be described by a linear, second-order model with the assumption that the mechanical properties of the knee were constant over these small displacements. The elasticity about the knee was influenced not only by the static joint angle but also by the perturbation intensity. Realistic models of the knee joint system should be modeled by a position-dependent, nonlinear system to remain valid over a large range of rotation.     

A study of shoulder motions as a control source for adolescents with C4 level SCI.

This study quantitatively examined and compared the shoulder motions of C4 level spinal cord injury (SCI), C5 level SCI, and able-bodied persons as a command source. The study was motivated by both the success of shoulder control in functional electrical stimulation (FES) systems designed for C5 level SCI people and the lack of quantitative information on the shoulder motion of persons with C4 level SCI. A dual-axis transducer was used to monitor the elevation/depression and protraction/retraction angles of each subject's shoulder while they performed three experimental sections which examined: the range of active shoulder motion; the ability to move incrementally to discrete positions with the aid of visual feedback; and the ability to hold discrete shoulder positions for an extended period without visual feedback. Results indicated that each group had the largest average shoulder displacements (abled = 23 degrees +/- 4 degrees, C5's = 14 degrees +/- 3 degrees, and C4's = 9 degrees +/- 3 degrees) while attempting to elevate and that on average the C4 group had the smallest range of active shoulder motion. No statistically significant differences between the groups were found in either the accuracy or stability of reaching discrete positions with the aid of visual feedback or in the accuracy of holding discrete shoulder positions for an extended period without visual feedback. The results suggest that within their limited range of motion the individuals with C4 level SCI retained shoulder control sufficient for use as an neuroprosthetic command interface.     

Conveying Tactile Feedback in Sensorized Hand Neuroprostheses Using a Biofidelic Model of Mechanotransduction

One approach to conveying tactile feedback from sensorized neural prostheses is to characterize the neural signals that would normally be produced in an intact limb and reproduce them through electrical stimulation of the residual peripheral nerves. Toward this end, we have developed a model that accurately replicates the neural activity evoked by any dynamic stimulus in the three types of mechanoreceptive afferents that innervate the glabrous skin of the hand. The model takes as input the position of the stimulus as a function of time, along with its first (velocity), second (acceleration), and third (jerk) derivatives. This input is filtered and passed through an integrate-and-fire mechanism to generate a train of spikes as output. The major conclusion of this study is that the timing of individual spikes evoked in mechanoreceptive fibers innervating the hand can be accurately predicted by this model. We discuss how this model can be integrated in a sensorized prosthesis and show that the activity in a population of simulated afferents conveys information about the location, timing, and magnitude of contact between the hand and an object.      

边宽尾墩体型对边墙区域水流水力特性的影响研究

本文采用模型实验和数值模拟方法对Y型宽尾墩+阶梯溢流坝+消力池一体化消能进行分析研究,Y型边宽尾墩在满足泄洪消能条件下,侧收缩角设置不佳,会出现水流飞溅出边墙的问题。通过实验结果,比较分析三种不同体型边宽尾墩的整体水流流态、墩后水翅、溢流坝面时均压强、消力池时均压强及流速分布发现,边宽尾墩侧收缩角分别为19.03°、17.07°、19.98°、13.80°和19.98°、7.0°时消力池内水流流态均较好,其消能效果较好。当边宽尾墩侧收缩角为19.98°、7.0°时,水流不飞溅出边墙。