Restoration of lateral hand grasp using natural sensors

Improved Signal Processing Methods for Velocity Selective Neural Recording Using Multi-Electrode Cuffs

This paper describes an improved system for obtaining velocity spectral information from electroneurogram recordings using multi-electrode cuffs (MECs). The starting point for this study is some recently published work that considers the limitations of conventional linear signal processing methods (`delay-and-add') with and without additive noise. By contrast to earlier linear methods, the present paper adopts a fundamentally non-linear velocity classification approach based on a type of artificial neural network (ANN). The new method provides a unified approach to the solution of the two main problems of the earlier delay-and-add technique, i.e., a damaging decline in both velocity selectivity and velocity resolution at high velocities. The new method can operate in real-time, is shown to be robust in the presence of noise and also to be relatively insensitive to the form of the action potential waveforms being classified.

Comparative analysis of transverse intrafascicular multichannel, longitudinal intrafascicular and multipolar cuff electrodes for the selective stimulation of nerve fascicles

The selection of a suitable nerve electrode for neuroprosthetic applications implies a trade-off between invasiveness and selectivity, wherein the ultimate goal is achieving the highest selectivity for a high number of nerve fascicles by the least invasiveness and potential damage to the nerve. The transverse intrafascicular multichannel electrode (TIME) is intended to be transversally inserted into the peripheral nerve and to be useful to selectively activate subsets of axons in different fascicles within the same nerve. We present a comparative study of TIME, LIFE and multipolar cuff electrodes for the selective stimulation of small nerves. The electrodes were implanted on the rat sciatic nerve, and the activation of gastrocnemius, plantar and tibialis anterior muscles was recorded by EMG signals. Thus, the study allowed us to ascertain the selectivity of stimulation at the interfascicular and also at the intrafascicular level. The results of this study indicate that (1) intrafascicular electrodes (LIFE and TIME) provide excitation circumscribed to the implanted fascicle, whereas extraneural electrodes (cuffs) predominantly excite nerve fascicles located superficially; (2) the minimum threshold for muscle activation with TIME and LIFE was significantly lower than with cuff electrodes; (3) TIME allowed us to selectively activate the three tested muscles when stimulating through different active sites of one device, both at inter- and intrafascicular levels, whereas selective activation using multipolar cuff (with a longitudinal tripolar stimulation configuration) was only possible for two muscles, at the interfascicular level, and LIFE did not activate selectively more than one muscle in the implanted nerve fascicle.

A summary of the theory of velocity selective neural recording

This paper describes improvements to the technique of velocity selective recording (VSR) in which multiple neural signals are matched and summed to identify excited axon populations in terms of velocity. This form of recording has been termed intrinsic velocity selective recording (IVSR). The signals are acquired using a multi-electrode cuff (MEC) which is now available as a component for use in implantable neuroprostheses. The improvements outlined in the paper involve the use of bandpass filters at the output of the system which allows a higher level of selectivity to be obtained than is possible using IVSR.

Effects of consecutive slips in nerve signals recorded by implanted cuff electrode

Using an anaesthetized cat's central footpad pressed against an object as the model of a paralyzed human hand, a nerve signal recording system was developed to measure the effect of a group of consecutive slips between the footpad and the object. Electroneurographic (ENG) activity was recorded using a cuff electrode implanted around the tibial nerve. The relationship between the recorded nerve signals during consecutive slips was investigated. The analyzed results showed that the amplitude of the ENG signal corresponding to the first slip was significantly greater than subsequent slips. It was also shown that there was no significant difference in the amplitude of the ENG signal in subsequent slips. When the slip signal is used as a feedback and control signal for FES, two different thresholds or scaling factors should be applied for consecutive slips.

Model-based evaluation of the short-circuited tripolar cuff configuration

Recordings of neural information for use as feedback in functional electrical stimulation are often contaminated with interfering signals from muscles and from stimulus pulses. The cuff electrode used for the neural recording can be optimized to improve the S/I ratio. In this work, we evaluate a model of both the nerve signal and the interfering signals recorded by a cuff, and subsequently use this model to study the signal to interference ratio of different cuff designs and to evaluate a recently introduced short-circuited tripolar cuff configuration. The results of the model showed good agreement with results from measurements in rabbits and confirmed the superior performance of the short-circuited tripolar configuration as compared with the traditionally used tripolar configuration.

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 degrees (a total excursion of 20 degrees) 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 degrees. 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.

Indications and future directions for upper limb neuroprostheses in tetraplegic patients: a review

The development of the upper extremity neuroprosthesis has been a challenging and rewarding contribution to the management of the SCI patient. The authors' experience and that of their clinical trial teams has verified that this technology is a strong alternative to conventional reconstruction and conservative management. In the future, even more powerful tools will emerge from the laboratory as these devices and collaborative surgical procedures evolve.

Neuroprostheses for the upper extremity

Functional electrical stimulation (FES) neuroprostheses can be used to replace lost motor and sensory function in persons with neurological disorders. FES technology has subsequently been shown effective and safe in restoring hand function in adults with spinal cord injury. The freehand system consists of an implanted receiver-stimulator, an external shoulder position sensor, and an external control unit. Commands are originated by voluntary movement of the contralateral shoulder and are measured by the sensor. There are several types of electrodes: epimysial, intramuscular, nerve cuff, and intraneural. Neuroprostheses are recommended within the context of all available reconstructive options for the upper limbs. Voluntary tendon transfers are the first choice. The clinical outcomes as measured by improvement on scales of impairment, activities of daily living, and satisfaction are rewarding. The next step in improvement of the motor function of person with spinal cord injury will be the addition of a controllable second upper extremity and the elimination of additional external hardware.

Effect of initial joint position on nerve-cuff recordings of muscle afferents in rabbits

The objective was to characterize nerve-cuff recordings of muscle afferents to joint rotation over a large part of the physiological joint range. This information is needed to develop control strategies for functional electrical stimulation (FES) systems using muscle afferent signals for sensory feedback. Five acute rabbit experiments were performed. Tripolar cuff electrodes were implanted around the tibial and peroneal divisions of the sciatic nerve in the rabbit's left leg. The electroneurograms (ENG) were recorded during passive ankle rotation, using a ramp-and-hold profile starting at seven different joint positions (excursion = 5 degrees, velocity = 10 degrees/s, initial positions 60 degrees, 70 degrees, 80 degrees, 90 degrees, 100%, 110 , and 120 ). The amplitude of the afferent activity was dependent on the initial joint position. The steady-state sensitivity of both nerve responses increased with increasing joint flexion, whereas the dynamic sensitivity increased initially but then decreased. The results indicate that recordings of the muscle afferents may provide reliable information over only a part of the physiological joint range. Despite this limitation, muscle afferent activity may be useful for motion feedback if the movement to be controlled is within a narrow joint range such as postural sway.

Effect of agonist-antagonist electrical stimulation on muscle afferent recordings in anesthetized rabbits

Objective. Sensory feedback extracted from muscle afferents is an approach to achieve closed-loop control of paralyzed muscles using functional electrical stimulation (FES). The objective of the present study was to characterize the effect of agonist-antagonist electrical stimulation on nerve cuff recordings of muscle afferent activity. Methods. Cuff electrodes were implanted around the tibial and peroneal nerve branches in five acute rabbit experiments. Two wires were implanted in each of the tibialis anterior (TA) and the lateral gastrocnemius (LG) muscles to obtain bipolar, intramuscular stimulation. Electroneurograms (ENG) were recorded during trapezoidal rotations of the ankle joint and compared during periods (25%, 50% and 100% of maximal force) with and without electrical stimulation of the muscle. Results. The activity from a stretched and electrically stimulated muscle showed the same pattern as the recordings from a matched nonstimulated muscle. The background afferent activity increased with increasing level of muscle stimulation. The static and dynamic sensitivities were not found to be different, except in one case (peroneal nerve at 100% TA recruitment). Discussion. The main contribution to the tibial activity was believed to originate from muscle afferents in nonstimulated, synergist muscles. The main contribution to the peroneal activity was believed to be from muscle afferents within the muscle being stimulated. It was suggested that the increased background activity could be attributed to the increased activity of the Golgi tendon organs. Conclusions. Sensory information about joint flexion and joint extension are preserved in muscle afferent recordings from electrically activated muscles at low and intermediate stimulation levels, but it still has to be shown whether muscle afferent information can be useful as sensory feedback in FES control.

Neuro-fuzzy extraction of angular information from muscle afferents for ankle control during standing in paraplegic subjects: an animal model

This paper is part of a project whose aim is the implementation of closed-loop control of ankle angular position during functional electrical stimulation (FES) assisted standing in paraplegic subjects using natural sensory information. In this paper, a neural fuzzy (NF) model is implemented to extract angular position information from the electroneurographic signals recorded from muscle afferents using cuff electrodes in an animal model. The NF model, named dynamic nonsingleton fuzzy logic system is a Mamdani-like fuzzy system, implemented in the framework of recurrent neural networks. The fuzzification procedure implemented was the nonsingleton technique which has been shown in previous works to be able to take into account the uncertainty in the data. The proposed algorithm was tested in different situations and was able to predict reasonably well the ankle angular trajectories especially for small excursions (as during standing) and when the stimulation sites are far from the registration sites. This suggests it may be possible to use activity from muscle afferents recorded with cuff electrodes for FES closed-loop control of ankle position during quite standing.

Biomechatronics--assisting the impaired motor system

Biomechatronics concerns the interdisciplinary field of interaction with the human neuromuscular-skeletal system with the objective to assist impaired human motor control. In this field technology is developed that integrates neuroscience, robotics, interface and sensor technology, dynamic systems and control theory. The primary issue in this field concerns the concepts of assisting impaired human motor function. The secondary, derived, issue concerns possible methods of interfacing with the human body at all hierarchical levels of the human motor system. The application of motor assist systems may serve several goals: it can take over part of the affected motor control, enable the physiological motor system to perform the desired function or aid in training the impaired physiological system. The progress in these issues are reviewed and their potential implications for assistance of the impaired human motor system are discussed.

Neuroprosthetic applications of electrical stimulation

Neural prostheses are a developing technology that use electrical activation of the nervous system to restore function to individuals with neurological impairment. Neural prostheses function by electrical initiation of action potentials in nerve fibers that carry the signal to an endpoint where chemical neurotransmitters are released, either to affect an end organ or another neuron. Thus, in principle, any end organ under neural control is a candidate for neural prosthetic control. Applications have included stimulation in both the sensory and motor systems and range in scope from experimental trials with single individuals to commercially available devices. Outcomes of motor system neural prostheses include restoration of hand grasp and release in quadriplegia, restoration of standing and stepping in paraplegia, restoration of bladder function (continence, micturition) following spinal cord injury, and electrophrenic respiration in high-level quadriplegia. Neural prostheses restore function and provide greater independence to individuals with disability.

Measurement of the performance of nerve cuff electrodes for recording

New designs of cuff electrodes for the recording of signals from peripheral nerves are typically tested in acute animal experiments before long-term evaluation takes place. A reproducible, cost-effective and fast method is presented for evaluating cuff electrodes with respect to signal amplitude, noise rejection, and, in some cases, selectivity, as an alternative to acute in vivo experiments. Comparisons with a computer model and with signals obtained from rabbit tibial nerve give good agreement with the new method. It is shown that an imperfect closure of the cuff around the nerve can easily lead to more than 50% loss of the signal amplitude. Noise from sources external to the cuff is not significantly affected by the closing mechanism, but is strongly reduced by a tripolar cuff configuration as compared with a monopolar one (reduction factor 2.8 to 58, mean = 6.5, n = 6). In dual-channel cuffs, cross-talk is below 1.2% indicating a very high selectivity.

Control of FES thumb force using slip information obtained from the cutaneous electroneurogram in quadriplegic man

A tetraplegic volunteer was implanted with percutaneous intramuscular electrodes in hand and forearm muscles. Furthermore, a sensory nerve cuff electrode was implanted on the volar digital nerve to the radial side of the index finger branching off the median nerve. In laboratory experiments a stimulation system was used to produce a lateral grasp (key grip) while the neural activity was recorded with the cuff electrode. The nerve signal contained information that could be used to detect the occurrence of slips and further to increase stimulation intensity to the thumb flexor/adductor muscles to stop the slip. Thereby the system provided a grasp that could catch an object if it started to slip due to, e.g., decreasing muscle force or changes in load forces tangential to the surface of the object. This method enabled an automatic adjustment of the stimulation intensity to the lowest possible level without loosing the grip and without any prior knowledge about the strength of the muscles and the weight and surface texture of the object.

Characterization of signals and noise rejection with bipolar longitudinal intrafascicular electrodes

Longitudinal intrafascicular electrodes (LIFE's) are fine electrodes threaded into the extracellular space between axons in peripheral nerves or spinal roots. We are developing these electrodes for application in functional electrical stimulation and in basic physiology. An area of concern in chronic recording application of LIFE's is the possibility of electromyogram and other external noise sources masking the recorded neural signals. We characterized neural signals recorded by LIFE's and confirmed by three independent methods that increasing interelectrode spacing for bipolar LIFE's increases signal amplitude. The spectrum of neural signal from bipolar and monopolar LIFE lies between 300 Hz and 10 kHz. The amplitude of the spectrum increases with increasing interelectrode spacing, although the distribution is not affected. Single unit analysis of LIFE recordings show that they record selectively from units closest to the electrode active site. Units with conduction velocities ranging from 50-120 m/s were identified. Extraneural noise, as stimulus artifact or electromyogram, is much reduced with bipolar LIFE recording, as compared to monopolar recordings. Relative improvement in neural signal to extraneural noise increases with interelectrode spacing up to about 2 mm. Since there is no further improvement beyond 2 mm, we conclude that the preferred interelectrode spacing for bipolar LIFE's is 2 mm.