Past, present, future: a review on visual prostheses.

This review presents the historic background, the present status and the future prospects of several visual implant types such as retinal implants, as well as optic nerve and thalamus stimulation, and stimulation of the visual cortex. Already achieved milestones, such as improved object recognition and letter reading, give hope that a breakthrough can be achieved in the coming years. Furthermore, clinical results from recent studies are presented in order to describe the obstacles that have to be overcome in the future. Currently, inner eye implants placed within the retina are the preferred option for electrical stimulation of the optic pathway. However, the limited number of stimulating electrodes and the non-focal activation of target neurons are still hindering the generation of percepts of higher quality. In the future, optogenetic approaches may be used to activate retinal neurons with a high temporal and spatial accuracy. The main challenge for all types of visual implants will be to utilize as many parts of the remaining optic pathway as possible and to focally activate functionally different pathways.

Modeling the response of ON and OFF retinal bipolar cells during electric stimulation.

Retinal implants allowing blind people suffering from diseases like retinitis pigmentosa and macular degeneration to regain rudimentary vision are struggling with several obstacles. One of the main problems during external electric stimulation is the co-activation of the ON and OFF pathways which results in mutual impairment. In this study the response of ON and OFF cone retinal bipolar cells during extracellular electric stimulation from the subretinal space was examined. To gain deeper insight into the behavior of these cells sustained L-type and transient T-type calcium channels were integrated in the synaptic terminals of reconstructed 3D morphologies of ON and OFF cone bipolar cells. Intracellular calcium concentration in the synaptic regions of the model neurons was investigated as well since calcium influx is a crucial parameter for cell-to-cell activity between bipolar cells and retinal ganglion cells. It was shown that monophasic stimulation results in significant different calcium concentrations in the synaptic terminals of ON and OFF bipolar cells. Intracellular calcium increased to values up to fourfold higher in the OFF bipolar model neuron in comparison to the ON bipolar cell. Furthermore, geometric properties strongly influence the activation of bipolar cells. Monophasic, biphasic, single and repetitive pulses with similar lengths, amplitudes and polarities were applied to the two model neurons.

Influence of the sodium channel band on retinal ganglion cell excitation during electric stimulation--a modeling study.

Electric stimulation using retinal implants allows blind people to re-experience a rudimentary kind of vision. The elicited percepts or so called 'phosphenes' are highly inconstant and therefore do not restore vision properly. The better knowledge of how retinal neurons, especially retinal ganglion cells, respond to electric stimulation will help to develop more sophisticated stimulation strategies. Special anatomic and physiologic properties like a band of highly dense sodium channels in retinal ganglion cells may help to achieve a focal activation of target cells and as a result better restoration of vision. A portion of retinal ganglion cell axons, about 40µm from the soma and between 25 and 40µm in length, shows a specific biophysical property. Electrode locations close to a band of highly dense sodium channels which were identified immunochemically show lowest thresholds during electric stimulation. The (modeled) thresholds for this kind of structure result in lowest thresholds as well. The influence on the location where action potentials are generated within the axon is far reaching. When a stimulating electrode is positioned far outside the actual band region the site of spike initiation still remains within the sodium channel band. These findings suggest to further examine the key mechanisms of activation for retinal ganglion cells because focal activation without influencing passing axons of neurons located far away can improve the outcome of electric stimulation and therefore the development of retinal implants.

Morphometric classification and spatial organization of spiral ganglion neurons in the human cochlea: consequences for single fiber response to electrical stimulation.

The unique, unmyelinated perikarya of spiral ganglion cells (SGCs) in the human cochlea are often arranged in functional units covered by common satellite glial cells. This micro anatomical peculiarity presents a crucial barrier for an action potential (AP) travelling from the sensory receptors to the brain. Confocal microscopy was used to acquire systematically volumetric data on perikarya and corresponding nuclei in their full dimension along the cochlea of two individuals. Four populations of SGCs within the human inner ear of two different specimens were identified using agglomerative hierarchical clustering, contrary to the present distinction of two groups of SGCs. Furthermore, we found evidence of a spatial arrangement of perikarya and their accordant nuclei along the cochlea spiral. In this arrangement, the most uniform sizes of cell bodies are located in the middle turn, which represents the majority of phonational frequencies. Since single-cell recordings from other mammalians may not be representative to humans and human SGCs are not accessible for physiological measurements, computer simulation has been used to quantify the effect of varying soma size on single neuron response to electrical micro stimulation. Results show that temporal parameters of the spiking pattern are affected by the size of the cell body. Cathodic stimulation was found to induce stronger variations of spikes while also leading to the lowest thresholds and longest latencies. Therefore, anodic stimulation leads to a more uniform excitation profile among SGCs with different cell body size.

Strength-duration relationship for intra- versus extracellular stimulation with microelectrodes.

Chronaxie, a historically introduced excitability time parameter for electrical stimulation, has been assumed to be closely related to the time constant of the cell membrane. Therefore, it is perplexing that significantly larger chronaxies have been found for intracellular than for extracellular stimulation. Using compartmental model analysis, this controversy is explained on the basis that extracellular stimulation also generates hyperpolarized regions of the cell membrane hindering a steady excitation as seen in the intracellular case. The largest inside/outside chronaxie ratio for microelectrode stimulation is found in close vicinity of the cell. In the case of monophasic cathodic stimulation, the length of the primarily excited zone which is situated between the hyperpolarized regions increases with electrode-cell distance. For distant electrodes this results in an excitation process comparable to the temporal behavior of intracellular stimulation. Chronaxie also varies along the neural axis, being small for electrode positions at the nodes of Ranvier and axon initial segment and larger at the soma and dendrites. As spike initiation site can change for short and long pulses, in some cases strength-duration curves have a bimodal shape, and thus, they deviate from a classical monotonic curve as described by the formulas of Lapicque or Weiss.

Assessment criteria for evaluating the stability and position of the centre of gravity on a balance training platform: a simulation with Matlab®.

The anterior-posterior and medial-lateral tipping values of a training platform within the scope of a balance training activity should be evaluated in regard to position and the changing position of the centre of gravity (COG). Data streams are simulated and modulated with the help of the signal processing programme Matlab®. These data streams are evaluated using existing balance formulas and a specially developed formula. Active Balance Index referring to the zero point (ABI(0)) allows an assertion about the magnitude of the mean tipping angle with the magnitude origin in the centre of the training platform. Active Balance Index referring to the arithmetical mean COG (ABI(mean)) enables an assertion about the mean tipping angle with the magnitude origin in the COG mean position. The deviation of the mean COG from absolute balance which is projected onto the platform is calculated by the Active Balance Ratio (ABR). The Active Balance Index (ABI) in combination with the ABR seems to be an adequate parameter for the evaluation of COG stability on a training platform.

Design and development of an automatic data acquisition system for a balance study using a smartcard system.

For measurement value logging of board angle values during balance training, it is necessary to develop a measurement system. This study will provide data for a balance study using the smartcard. The data acquisition comes automatically. An individually training plan for each proband is necessary. To store the proband identification a smartcard with an I2C data bus protocol and an E2PROM memory system is used. For reading the smartcard data a smartcard reader is connected via universal serial bus (USB) to a notebook. The data acquisition and smartcard read programme is designed with Microsoft® Visual C#. A training plan file contains the individual training plan for each proband. The data of the test persons are saved in a proband directory. Each event is automatically saved as a log-file for the exact documentation. This system makes study development easy and time-saving.

Which elements of the mammalian central nervous system are excited by low current stimulation with microelectrodes?

Low current cortex stimulation produces a sparse and distributed set of activated cells often with distances of several hundred micrometers between cell bodies and the microelectrode. A modeling study based on recently measured densities of high threshold sodium channels Nav1.2 in dendrites and soma and low threshold sodium channels Nav1.6 in the axon shall identify spike initiation sites including a discussion on dendritic spikes. Varying excitability along the neural axis has been observed while studying different electrode positions and configurations. Although the axon initial segment (AIS) and nodes of Ranvier are most excitable, many thin axons and dendrites which are likely to be close to the electrode in the densely packed cortical regions are also proper candidates for spike initiation sites. Cathodic threshold ratio for thin axons and dendrites is about 1:3, whereas 0.2 mum diameter axons passing the electrode tip in 10 mum distance can be activated by 100 mus pulses with 2.6 muA. Direct cathodic excitation of dendrites requires a minimum electrode-fiber distance, which increases with dendrite diameter. Therefore thin dendrites can profit from the stronger electrical field close to the electrode but low current stimulation cannot activate large diameter dendrites, contrary to the inverse recruitment order known from peripheral nerve stimulation. When local depolarization fails to generate a dendritic spike, stimulation is possible via intracellular current flow that initiates an action potential, for example 200 mum distant in the low threshold AIS or in certain cases at the distal dendrite ending. Beside these exceptions, spike initiation site for cathodic low current stimulation appears rather close to the electrode.

Human lumbar cord circuitries can be activated by extrinsic tonic input to generate locomotor-like activity.

We have demonstrated that non-patterned electrical stimulation of the lumbar cord can induce stepping-like activity in the lower limbs of complete spinal cord injured individuals. This result suggested the existence of a human lumbar locomotor pattern generator, which can convert a tonic input to a rhythmic motor output. We have studied the human lumbar cord in isolation from supraspinal input but under extrinsic tonic input delivered by spinal cord stimulation. Large-diameter afferents within the posterior roots are directly depolarized by the electrical stimulation. These afferents project to motoneurons as well as to lumbar interneurons involved in the motor control of lower limbs. Stimulation at 25-50 Hz can elicit rhythmic alternating flexion/extension movements of the lower limbs in supine individuals. Reducing the tonic input frequency to 5-15 Hz initiates lower limb extension. Epidural stimulation applied during manually assisted treadmill stepping in complete spinal cord injured persons immediately increases the central state of excitability of lumbar cord networks and enhances stepping-like functional motor outputs. Sustained, non-patterned tonic input via the posterior roots can activate human lumbar cord networks. Pattern generating configurations of these multifunctional circuitries can be set-up depending on the stimulation parameters and particularly on the input frequency.

Stepping-like movements in humans with complete spinal cord injury induced by epidural stimulation of the lumbar cord: electromyographic study of compound muscle action potentials.

STUDY DESIGN: It has been previously demonstrated that sustained nonpatterned electric stimulation of the posterior lumbar spinal cord from the epidural space can induce stepping-like movements in subjects with chronic, complete spinal cord injury. In the present paper, we explore physiologically related components of electromyographic (EMG) recordings during the induced stepping-like activity. OBJECTIVES: To examine mechanisms underlying the stepping-like movements activated by electrical epidural stimulation of posterior lumbar cord structures. MATERIALS AND METHODS: The study is based on the assessment of epidural stimulation to control spasticity by simultaneous recordings of the electromyographic activity of quadriceps, hamstrings, tibialis anterior, and triceps surae. We examined induced muscle responses to stimulation frequencies of 2.2-50 Hz in 10 subjects classified as having a motor complete spinal cord injury (ASIA A and B). We evaluated stimulus-triggered time windows 50 ms in length from the original EMG traces. Stimulus-evoked compound muscle action potentials (CMAPs) were analyzed with reference to latency, amplitude, and shape. RESULTS: Epidural stimulation of the posterior lumbosacral cord recruited lower limb muscles in a segmental-selective way, which was characteristic for posterior root stimulation. A 2.2 Hz stimulation elicited stimulus-coupled CMAPs of short latency which were approximately half that of phasic stretch reflex latencies for the respective muscle groups. EMG amplitudes were stimulus-strength dependent. Stimulation at 5-15 and 25-50 Hz elicited sustained tonic and rhythmic activity, respectively, and initiated lower limb extension or stepping-like movements representing different levels of muscle synergies. All EMG responses, even during burst-style phases were composed of separate stimulus-triggered CMAPs with characteristic amplitude modulations. During burst-style phases, a significant increase of CMAP latencies by about 10 ms was observed. CONCLUSION: The muscle activity evoked by epidural lumbar cord stimulation as described in the present study was initiated within the posterior roots. These posterior roots muscle reflex responses (PRMRRs) to 2.2 Hz stimulation were routed through monosynaptic pathways. Sustained stimulation at 5-50 Hz engaged central spinal PRMRR components. We propose that repeated volleys delivered to the lumbar cord via the posterior roots can effectively modify the central state of spinal circuits by temporarily combining them into functional units generating integrated motor behavior of sustained extension and rhythmic flexion/extension movements. This study opens the possibility for developing neuroprostheses for activation of inherent spinal networks involved in generating functional synergistic movements using a single electrode implanted in a localized and stable region.

Initiating extension of the lower limbs in subjects with complete spinal cord injury by epidural lumbar cord stimulation.

We provide evidence that the human spinal cord is able to respond to external afferent input and to generate a sustained extension of the lower extremities when isolated from brain control. The present study demonstrates that sustained, nonpatterned electrical stimulation of the lumbosacral cord--applied at a frequency in the range of 5-15 Hz and a strength above the thresholds for twitches in the thigh and leg muscles--can initiate and retain lower-limb extension in paraplegic subjects with a long history of complete spinal cord injury. We hypothesize that the induced extension is due to tonic input applied by the epidural stimulation to primary sensory afferents. The induced volleys elicit muscle twitches (posterior root muscle-reflex responses) at short and constant latency times and coactivate the configuration of the lumbosacral interneuronal network, presumably via collaterals of the primary sensory neurons and their connectivity with this network. We speculate that the volleys induced externally to the lumbosacral network at a frequency of 5-15 Hz initiate and retain an "extension pattern generator" organization. Once established, this organization would recruit a larger population of motor units in the hip and ankle extensor muscles as compared to the flexors, resulting in an extension movement of the lower limbs. In the electromyograms of the lower-limb muscle groups, such activity is reflected as a characteristic spatiotemporal pattern of compound motor-unit potentials.

Mechanisms of electrical stimulation with neural prostheses.

Individual electric and geometric characteristics of neural substructures can have surprising effects on artificially controlled neural signaling. A rule of thumb approved for the stimulation of long peripheral axons may not hold when the central nervous system is involved. This is demonstrated here with a comparison of results from the electrically stimulated cochlea, retina, and spinal cord. A generalized form of the activating function together with accurate modeling of the neural membrane dynamics are the tools to analyze the excitation mechanisms initiated by neural prostheses. Analysis is sometimes possible with a linear theory, in other cases, simulation of internal calcium concentration or ion channel current fluctuations is needed to see irregularities in spike trains. Spike initiation site can easily change within a single target neuron under constant stimulation conditions of a cochlear implant. Poor myelinization in the soma region of the human cochlear neurons causes firing characteristics different from any animal data. Retinal ganglion cells also generate propagating spikes within the dendritic tree. Bipolar cells in the retina are expected to respond with neurotransmitter release before a spike is generated in the ganglion cell, even when they are far away from the electrode. Epidural stimulation of the lumbar spinal cord predominantly stimulates large sensory axons in the dorsal roots which induce muscle reflex responses. Analysis with the generalized activating function, computer simulations of the nonlinear neural membrane behavior together with experimental and clinical data analysis enlighten our understanding of artificial firing patterns influenced by neural prostheses.

A model of the electrically excited human cochlear neuron. I. Contribution of neural substructures to the generation and propagation of spikes.

Differences in neural geometry and the fact that the soma of the human cochlear neuron typically is not myelinated are reasons for disagreements between single fiber recordings in animals and the neural code evoked in cochlear implant patients. We introduce a compartment model of the human cochlear neuron to study the excitation and propagation process of action potentials. The model can be used to predict (i) the points of spike generation, (ii) the time difference between stimulation and the arrival of a spike at the proximal end of the central axon, (iii) the vanishing of peripherally evoked spikes at the soma region under specific conditions, (iv) the influence of electrode positions on spiking behavior, and (v) consequences of the loss of the peripheral axon. Every subunit of the cochlear neuron is separately modeled. Ion channel dynamics are described by a modified Hodgkin--Huxley model. Influence of membrane noise is taken into account. Additionally, the generalized activating function is introduced as a tool to give an envision of the origin of spikes in the peripheral and in the central axon without any knowledge of the gating processes in the active membranes. Comparing the reactions of a human and cat cochlear neuron, we find differences in spiking behavior, e.g. peripherally and centrally evoked spikes arrive with a time difference of about 400 mus in man and 200 mus in cat.

A model of the electrically excited human cochlear neuron. II. Influence of the three-dimensional cochlear structure on neural excitability.

A simplified spiraled model of the human cochlea is developed from a cross sectional photograph. The potential distribution within this model cochlea is calculated with the finite element technique for an active scala tympani implant. The method in the companion article [Rattay et al., 2001] allows for simulation of the excitation process of selected elements of the cochlear nerve. The bony boundary has an insulating influence along every nerve fiber which shifts the stimulation condition from that of a homogeneous extracellular medium towards constant field stimulation: for a target neuron which is stimulated by a ring electrode positioned just below the peripheral end of the fiber the extracellular voltage profile is rather linear. About half of the cochlear neurons of a completely innervated cochlea are excited with monopolar stimulation at three-fold threshold intensity, whereas bipolar and especially quadrupolar stimulation focuses the excited region even for stronger stimuli. In contrast to single fiber experiments with cats, the long peripheral processes in human cochlear neurons cause first excitation in the periphery and, consequently, neurons with lost dendrite need higher stimuli.

Epidural electrical stimulation of posterior structures of the human lumbosacral cord: 2. quantitative analysis by computer modeling.

OBJECTIVES: Analysis of the computed recruitment order of an ensemble of ventral and dorsal root fibers should enlighten the relation between the position of a bipolar electrode and the observed order of muscle twitches. MATERIAL AND METHODS: Thresholds of selected spinal root fibers are investigated in a two step procedure. First the electric field generated by the electrodes is computed with the Finite Element Method. In the second step the calculated voltage profile along each target neuron is used as input data for a cable model. For every electrode position the electrical excitability is analyzed for 12 large diameter ventral and dorsal root fibers of the second and fourth lumbar and first sacral segment. The predictions of the neural responses of any target fiber are based on the activating function concept and on the more accurate computer simulations of the electrical behavior of all nodes and internodes in the vicinity of the electrode. RESULTS: For epidural dorsal lumbosacral spinal cord stimulation we found the following rules. (i) The recruitment order of the spinal roots is highly related to the cathode level. (ii) Dorsal root fibers have the lowest threshold values, ventral root fibers are more difficult to excite and dorsal columns are not excitable within the clinical range of 10 V. (iii) For a cathode close to the level of the spinal cord entry of a target fiber thresholds are lowest and spike initiation is expected at the border between cerebrospinal fluid and white matter; excitation of L4 roots is not possible with 210 micros/10 V pulses when cathode is more than 2.2 cm cranial to their entry level (1.5 cm for S1 roots; standard data). (iv) Cathodes positioned (essentially) below the entry level cause spike initiation close to the cathode, in a region where the fibers follow the descending course within the cerebospinal fluid. (v) At rather low stimulation voltage twitches are expected in all investigated lower limb muscles for cathodes below L5 spinal cord level. CONCLUSIONS: Our simulations demonstrate a strong relation between electrode position and the order of muscle twitches which is based on the segmental arrangement of innervation of lower limb muscles. The proposed strategy allows the identification of the position of the electrode relative to spinal cord segments.

Computer simulation of field distribution and excitation of denervated muscle fibers caused by surface electrodes.

In the course of this study, 2 submodels have been developed and combined, the 2-D finite element modeling of the electrical potential distribution in the human thigh and a Hodgkin and Huxley (HH) type model to calculate fiber excitation and action potential propagation. To determine the excitation of the target muscle fiber with the help of the activating function, the fiber's orientation within the muscle has to be known. The electric field along the fiber has to be calculated as a function of the applied electric current and the potential at the electrodes, respectively. The excitement of the muscle fibers varies across a wide range depending on the active and passive membrane parameters and the intracellular and extracellular mediums. Persisting denervation leads to a decay of muscle cells, and a partial substitution by fibroblasts occurs. The electrical activation of these tissues is more difficult, and biphasic stimulation pulses up to 200 ms in duration and 60-100 V in amplitude are needed to cause a contraction of the denervated muscle. An example shows the field distribution and the simulated activity in one representative muscle fiber of a well trained m. rectus femoris.

The basic mechanism for the electrical stimulation of the nervous system.

Neural signals can be generated or blocked by extracellular electrodes or magnetic coils. New results about artificial excitation are based on a compartmental model of a target neuron and its equivalent electrical network, as well as on the theory of the generalized activating function. The analysis shows that: (i) in most cases, the origin of artificial excitation is within the axon and the soma is much more difficult to excite; (ii) within the central nervous system, positive and negative threshold currents essentially depend on the position and orientation of the neurons relative to the applied electric field; (iii) in several cases, stimulation with positive currents is easier; and (iv) it should be possible to excite synaptic activity without the generation of propagating action potentials. Furthermore, the theory of the generalized activating function gives hints to understanding the blockage of neural activity.

Micromechanical models for the Brownian motion of hair cell stereocilia.

Brownian motion of the hairs (stereocilia) of amphibian hair cells has been shown in experiments to be in the range of some nm. Our models of the Brownian motion of coupled harmonic oscillators with mechanical properties of stereocilia lead to similar displacements. Computer simulation shows that stochastic fluctuations enhance the encoding of low level acoustic signals. Stochastic resonance lowers the detection threshold of auditory signals to amplitudes one order of magnitude lower than that of the Brownian motion.