Cochlear nucleus auditory prostheses.

Persons who lack an auditory nerve cannot benefit from cochlear implants, but a prosthesis utilizing an electrode array implanted on the surface of the cochlear nucleus can restore some hearing. Worldwide, more than 500 persons have received these "auditory brainstem implants," most commonly after removal of the tumors that occur with Type 2 Neurofibromatosis (NF2). Typically, the ABIs provide these individuals with improved speech perception when combined with lip-reading and useful perception of environmental sounds, but little open-set speech recognition. The feasibility of supplementing the array of surface electrodes with penetrating microstimulating electrodes has been investigated in animal studies, and 10 persons with NF2 have received implants that include a surface array and an array of penetrating microelectrodes. Their speech perception is not significantly better than that of the NF2 patients who have only the surface arrays, but the findings do validate the concept of intranuclear stimulation and suggest how such prostheses might be improved by modifying the microstimulating array and also by optimizing the sound processing strategies. Recent publications have described ABI patients with deafness of etiologies other than NF2 who have achieved open-set speech recognition. This suggests that the cochlear nuclei of the NF2 patients are damaged by the disease process or during surgical removal of the tumor.

"Safe" charge-injection waveforms for iridium oxide (AIROF) microelectrodes.

Use of anodic bias improves the charge-injection limits of activated iridium oxide (AIROF) microelectrodes. Asymmetric waveforms, in which the charge balancing anodic phase is delivered at a lower current density and longer pulse width, has been found to allow for higher values of anodic bias voltages, thus maximizing the AIROF charge-injection capacity. Limiting the voltage excursion of the AIROF below the value at which electrolysis of water occurs is essential to maintaining the long-term viability of implanted electrodes. However, maintaining the electrodes at an anodic bias state while keeping the electrode voltage within these electrochemically "safe" limits complicates the topology of the electronic driver circuitry. We present two possible driver topologies that use compliance-voltage limitation in combination with cathodic current modification.

Neural prostheses.

Assuming that neural regeneration after spinal cord injury (SCI) will eventually become a clinical reality, functional recovery will probably remain incomplete. Assistive devices will therefore continue to play an important role in rehabilitation. Neural prostheses (NPs) are assistive devices that restore functions lost as a result of neural damage. NPs electrically stimulate nerves and are either external or implanted devices. Surface stimulators for muscle exercise are now commonplace in rehabilitation clinics and many homes. Regarding implantable NPs, since 1963 over 40 000 have been implanted to restore hearing, bladder control and respiration. Epidural spinal cord stimulators and deep brain stimulators are routinely implanted to control pain, spasticity, tremor and rigidity. Implantable NPs have also been developed to restore limb movements using electrodes tunnelled under the skin to muscles and nerves. Spinal cord microstimulation (SC[mu]stim) is under study as an alternative way of restoring movement and bladder control. Improvement in bladder and bowel function is a high priority for many SCI people. Sacral root stimulation to elicit bladder contraction is the current NP approach, but this usually requires dorsal rhizotomies to reduce reflex contractions of the external urethral sphincter. It is possible that the spinal centres coordinating the bladder-sphincter synergy could be activated with SC[mu]stim. Given the large and growing number of NPs in use or development, it is surprising how little is known about their long-term interactions with the nervous system. Physiological research will play an important role in elucidating the mechanisms underlying these interactions.

Chronic microstimulation in the feline ventral cochlear nucleus: physiologic and histologic effects.

This study was conducted to help to establish the feasibility of a multi-channel auditory prosthesis based on microstimulation within the human ventral cochlear nucleus, and to define the range of stimulus parameters that can be used safely with such a device. We chronically implanted activated iridium microelectrodes into the feline ventral cochlear nucleus and, beginning 80-250 days after implantation, they were pulsed for 7 h/day, on up to 21 successive days. The stimulus was charge-balanced pulses whose amplitude was modulated by a simulated human voice. The pulse rate (250 Hz/electrode) and the maximum pulse amplitude were selected as those that are likely to provide a patient with useful auditory percepts. The changes in neuronal responses during the multi-day stimulation regimens were partitioned into long-lasting, stimulation-induced depression of neuronal excitability (SIDNE), and short-acting neuronal refractivity (SANR). Both SIDNE and SANR were quantified from the changes in the growth functions of the evoked potentials recorded in the inferior colliculus. All of the stimulation regimens that we tested induced measurable SIDNE and SANR. The combined effect of SIDNE and the superimposed SANR is to depress the neuronal response near threshold, and thereby, to depress the population response over the entire amplitude range of the stimulus pulses. SIDNE and SANR may cause the greatest degradation of the performance of a clinical device at the low end of the amplitude range, and this may represent an inherent limitation of this type of spatially localized, high-rate neuronal stimulation. We determined sets of stimulus parameters which preserved most of the dynamic range of the neuronal response, when using either long (150 micros/phase) or short (40 micros/phase) stimulus pulses. Increasing the amplitude of the stimulus was relatively ineffective as a means of increasing the dynamic range of neuronal response, since the greater stimulus amplitude induced more SIDNE. All of the pulsed and unpulsed electrode sites were examined histologically, and no neuronal changes attributable to the stimulation were detected. There was some aggregation of glial cells immediately adjacent to some of the electrodes that were pulsed with the short-duration pulses, and at the highest current densities.

Stability of the interface between neural tissue and chronically implanted intracortical microelectrodes.

The stability of the interface between neural tissue and chronically implanted microelectrodes is very important for obtaining reliable control signals for neuroprosthetic devices. Stability is also crucial for chronic microstimulation of the cerebral cortex. However, changes of the electrode-tissue interface can be caused by a variety of mechanisms. In the present study, intracortical microelectrode arrays were implanted into the pericruciate gyrus of cats and neural activities were recorded on a regular basis for several months. An algorithm based on cluster analysis and interspike interval analysis was developed to sort the extracellular action potentials into single units. We tracked these units based on their waveform and their response to somatic stimulation or stereotypical movements by the cats. Our results indicate that, after implantation, the electrode-tissue interface may change from day-to-day over the first 1-2 weeks, week-to-week for 1-2 months, and become quite stable thereafter. A stability index is proposed to quantify the stability of the electrode-tissue interface. The reasons for the pattern of changes are discussed.

Evolution and resolution of stimulation-induced axonal injury in peripheral nerve.

We describe the evolution of axonal injury following the induction of neural damage by electrical stimulation. The sciatic nerves of cats were stimulated continuously for 8 h with charge-balanced waveforms at high intensities, 50 Hz and 2100-4500 microA, using circumneural helical electrodes. Computer-assisted morphometric and ultrastructural studies indicate that many of the damaged fibers had not regenerated by 125 days after stimulation. Functional deficits were not observed in any of the animals, and most of the fibers appeared to be histologically normal at 125 days after stimulation. These findings indicate that there is relatively little late-onset injury associated with the stimulation. However, the slow, and possibly incomplete, recovery of the damaged axons emphasizes the importance of using stimulus protocols with adequate margins of safety.

Accessing the tonotopic organization of the ventral cochlear nucleus by intranuclear microstimulation.

This study is part of a program to develop an auditory prosthesis for the profoundly deaf, based on multichannel microstimulation in the cochlear nucleus. The functionality of such a device is dependent on its ability to access the tonotopic axis of the human ventral cochlear nucleus in an orderly fashion. In these studies, we utilized the homologies between the human and feline ventral cochlear nuclei and the known tonotopic organization of the central nucleus of the inferior colliculus (IC). In anesthetized cats, stimuli were delivered to three or four locations along the dorsal-to-ventral axis of the posteroventral cochlear nucleus (PVCN), and for each stimulus location, we recorded the multiunit neuronal activity and the field potentials at 20 or more locations along the dorsolateral-ventromedial (tonotopic) axis of the IC. The current source-sink density (CSD), which delimits regions of neuronal activity, was computed from the sequence of field potentials recorded along this axis. The multiunit activity and the CSD analysis both showed that the tonotopic organization of the PVCN can be accessed in an orderly manner by intranuclear microstimulation in several regions of the PVCN, using the range of stimulus pulse amplitudes that have been shown in previous studies to be noninjurious during prolonged intranuclear microstimulation via chronically implanted microelectrodes. We discuss the applicability of these findings to the design of clinical auditory prostheses for implantation into the human cochlear nucleus.

A characterization of the effects on neuronal excitability due to prolonged microstimulation with chronically implanted microelectrodes.

Localized, long-lasting stimulation-induced depression of neuronal excitability (SIDNE) is a consequence of prolonged, high-frequency microstimulation in the central nervous system (CNS). It represents a persisting refractory state in the neurons and axons near the stimulating microelectrode, that occurs in the absence of histologically detectable tissue injury. It does not involve a change in synaptic efficacy and, in this respect, it differs from the more familiar phenomenon of long-term depression (LTD). Although SIDNE is ultimately reversible (after several days), it must be taken into account in the design of neural prostheses based on microstimulation in the central nervous system and in animal studies that require prolonged microstimulation in the CNS. In this study, we have characterized the phenomenon, using as the paradigm, iridium microelectrodes implanted chronically in the cat's posteroventral cochlear nucleus. Although the SIDNE may persist for several days after the end of the stimulation protocol, it does not become more severe from day to day when the stimulation protocol is repeated on successive days. The severity of the SIDNE is strongly dependent upon both the instantaneous frequency and the duty cycle of the electrical stimulation. The character of the SIDNE, including its localization to the immediate vicinity of the stimulating microelectrodes, suggests that the phenomenon is a direct consequence of the prolonged electrical excitation of the neurons close to the microelectrode. The problem of designing microstimulation systems that allow high-frequency stimulation of a neural substrate, while minimizing SIDNE are discussed.

A quantitative computer-assisted morphometric analysis of stimulation-induced injury to myelinated fibers in a peripheral nerve.

We describe a computer-assisted morphometric procedure for quantifying acute axonal injury induced in peripheral nerves by prolonged electrical stimulation. The procedure is a two-phase process, with the image analysis implemented via a commercial image analysis program, followed by an automated editing of the morphometric parameters of each object identified by the image analysis software. Both phases are implemented on IBM-compatible personal computers. The custom software counts the number of fibers undergoing early axonal degeneration, using a two-category classification scheme based on the range of myelin cross-sectional area and axonal cross-sectional areas of normal (unstimulated) nerves. When the damaged fibers are counted using this procedure, the correlation between the normalized amplitude of the electric stimulus and the number of degenerating fibers is the same as when the analysis is performed by an experienced histopathologist (R = 0.87) and carries the advantage of being entirely objective. The correlation was higher with a two-category classification (damage/no damage) than when the severity of the damage to each axon was weighted according to the amount of axonal shrinkage. We determined that axons 3.5-9 microm in diameter are the most vulnerable to injury from the electrical stimulation. This has certain implications regarding the mechanism underlying this type of injury.

Threshold-distance measures from electrical stimulation of human brainstem.

Threshold current levels for electrical stimulation of a single human brainstem via an auditory prosthesis are compared with postmortem measures of the distance between the electrode and stimulated structures. The results compare well with the summary of threshold-distance measures from animal experiments compiled by Ranck. The correspondence between the human and animal data gives confidence that the extent of current spread (distance to stimulable neural units) can be well estimated from the current level at threshold for 200 microseconds/phase biphasic pulses. This is of particular interest in electrical stimulation of the human central nervous system, where localization of stimulation is of paramount importance.

Relationship between stimulus amplitude, stimulus frequency and neural damage during electrical stimulation of sciatic nerve of cat.

The relation is investigated between stimulus frequency, stimulus pulse amplitude and the neural damage induced by continuous stimulation of the cat's sciatic nerve. The chronically implanted electrodes were pulsed continuously and the effects of the electrical stimulation were quantified as the amount of early axonal degeneration (EAD) present in the nerves seven days after the continuous stimulation. The primary effect of stimulating at 100 Hz rather than 50 Hz was to cause an increase in the slope of the plot of the amount of EAD versus stimulus amplitude, but the threshold stimulus for the induction of EAD also was slightly lower. There was a small amount of EAD in three of the nerves stimulated at 20 Hz, but there was no detectable correlation between the amount of EAD and the stimulus amplitude. This suggests that continuous electrical stimulation of peripheral nerves at a low frequency induce little or no neural damage, even if the stimulus amplitude is very high. A preliminary presentation of the results has been made elsewhere (Agnew et al., 1993).

Stimulus parameters affecting tissue injury during microstimulation in the cochlear nucleus of the cat.

We investigated the effects of continuous microstimulation in the cats' posteroventral cochlear nucleus, using chronically implanted activated iridium microelectrodes. We examined 51 electrode sites (39 pulsed sites, and 12 unpulsed sites). Seven hours of continuous stimulation at 500 Hz often produced tissue injury near the tips of the pulsed microelectrodes. The damage took the form of a region of vacuolated tissue extending 200 microns or more from the site of the electrode tip. Electron microscope studies showed the vacuoles to be severely edematous segments of myelinated axons. The statistical correlation between the amount of damaged tissue and the charge per phase was large and highly significant (P < 0.0001). When the electrodes were pulsed for 7 h at 500 Hz with charge-balanced biphasic pulse pairs, the threshold for the damage was approximately 3 nC/phase. The damage threshold was not appreciably lower than the stimulation protocol was extended to 35 h (7 h/day for 5 days). In contrast, the threshold for exciting neurons near the microelectrode is approximately 1 nC/phase, as determined by the evoked response recorded in the inferior colliculus. There was little correlation between the severity of the tissue damage and the geometric charge density at the surface of the electrodes, between the damage and amplitude of the cathodic phase of the voltage transient induced across the stimulating electrodes by the stimulus current pulses, or between the damage and the stimulus pulse duration.

MK-801 protects against neuronal injury induced by electrical stimulation.

The ability of MK-801, a non-competitive N-methyl-D-aspartate receptor antagonist, to protect neurons in the cerebral cortex from injury induced by prolonged electrical stimulation was assessed in cats. Platinum disc electrodes 8.0 mm in diameter and with a surface area of 0.5 cm2 were implanted in the subdural space over the parietal cortex. Ten days after implantation of the electrodes, all animals received continuous stimulation for 7 h using charge-balanced, cathodic-first, controlled current pulses with a charge density of 20 microC/cm2 and a charge/phase of 10 microC/phase. They received either no MK-801, or 0.33 or 5.0 mg/kg (i.v.) administered intravenously, just before the start of the stimulation. Immediately following the stimulation, the animals were perfused and the cerebral cortex examined by light microscopy at eight sites beneath the electrodes. Neuronal damage in the form of shrunken, hyperchromic neurons and perineuronal halos was present only beneath the stimulating electrodes; damage was moderate to severe in stimulated animals that had not received MK-801, slight in animals receiving 0.33 mg/kg, and none to slight in animals receiving 5.0 mg/kg. These results indicate that MK-801, in an apparently dose-dependent fashion, provides substantial but not complete protection against neuronal injury induced by prolonged electrical stimulation. Thus prolonged electrical stimulation can be added to the list of neuropathologic conditions which involve glutamate-induced excitotoxic damage via the N-methyl-D-aspartate receptor. The results also support the hypothesis of neuronal hyperactivity as a principal cause of electrically-induced injury in the central nervous system. The implications for design of protocols for functional electrical stimulation are discussed.

Stimulation with chronically implanted microelectrodes in the cochlear nucleus of the cat: histologic and physiologic effects.

The effects of several hours of continuous electrical stimulation in the cats' cochlear nucleus with chronically implanted activated iridium microelectrodes was investigated from the changes in the evoked response near the inferior colliculus and also by histologic evaluation of the stimulated tissue. The stimulating microelectrodes had geometric surface areas of 75-500 microns2. They were pulsed continuously for 4 h, at a pulse repetition rate of 200 Hz, using charge-balanced pulse pairs. The charge per phase was 1.8 or 3.6 nC/ph. The animals were sacrificed for histologic evaluation 2 h, or several days later. The only remarkable histologic change resulting from the 4 h of stimulation was some aggregation of lymphocytes at the site of stimulation. However, depression of the electrical excitability of neurons near the sites often persisted for several days after 4 h of stimulation at 3.6 nC/phase. The charge per phase of the stimulus pulse pair was correlated strongly with the depression of excitability, and there was a weaker correlation between the depression and the amplitude of the first phase of voltage transient induced across the electrode-tissue interface. The charge density, calculated from the geometric surface area of the stimulating electrodes, was poorly correlated with the severity of the depression. The findings suggest a means of detecting impending stimulation-induced neural damage while it is still reversible.

Damage in peripheral nerve from continuous electrical stimulation: comparison of two stimulus waveforms.

The propensity for two types of charge-balanced stimulus waveforms to induce injury during eight hours of continuous electrical stimulation of the cat sciatic nerve was investigated. One waveform was a biphasic, controlled-current pulse pair, each phase 50 microseconds in duration, with no delay between the phases ('short pulse', selected to excite primarily large axons), whereas in the second type each phase was 100 microseconds in duration, with a 400 microsecond delay between the phases (selected to excite axons of a broader spectrum of diameters). The sciatic nerve was examined for early axonal degeneration (EAD) seven days after the session of continuous stimulation. With both waveforms, the threshold stimulus current for axonal injury was greater than the current required to excite all of the nerve's large axons. The correlation between simple stimulus parameters and the amount of EAD was poor, especially with the 'short pulse' waveform, probably due to variability between animals. When the stimulus was normalised with respect to the current required to fully recruit the large axons, a good association between damage and stimulus amplitude emerged. The damage threshold was higher for the 'short pulse' waveform. The implications for clinical protocols are discussed.

Charge density and charge per phase as cofactors in neural injury induced by electrical stimulation.

The possibility of neural injury during prolonged electrical stimulation of the brain imposes some constraints on the use of this technique for therapeutic and experimental applications. Stimulating electrodes of various sizes were used to investigate the interactions of two stimulus parameters, charge density and charge per phase, in determining the threshold of neural injury induced by electrical stimulation. Platinum electrodes ranging in size from 0.002 to 0.5 cm2 were implanted over the parietal cortex of adult cats. Penetrating microelectrodes fabricated from iridium, with surface areas of 65 +/- 3 x 10(-6) cm2 were inserted into the parietal cortex. Ten days after implantation, the electrodes were pulsed continuously for 7h using charge balanced, current regulated, symmetric pulse pairs, 400 microseconds per phase in duration, at a repetition rate of 50 Hz. The animals were perfused immediately after the stimulation for histologic evaluation of the brain tissue subjacent to the electrode sites. The results show that charge density (as measured at the surface of the stimulating electrode), and charge per phase, interact in a synergistic manner to determine the threshold of stimulation-induced neural injury. This interaction occurs over a wide range of both parameters; for charge density from at least 10 to 800 microC/cm2 and, for charge per phase, from at least 0.05 to 5.0 microC per phase. The significance of these findings in elucidating the mechanisms underlying stimulation-induced injury is discussed.

Local anaesthetic block protects against electrically-induced damage in peripheral nerve.

This study is one of a series addressing the mechanisms involved in the production of neural damage caused by continuous, prolonged electrical stimulation of peripheral nerve. It has been previously shown that sustained, high frequency electrical stimulation of the cat's peroneal nerve may cause irreversible neural damage in the form of axonal degeneration of the large myelinated fibres. In this study we demonstrate that blocking the action potentials on most of the nerve fibres with local anaesthetics (10% procaine or 2% lidocaine) almost completely prevents the axonal degeneration. The abolition of axonal injury by local anaesthetic block strongly suggests that the electrically-induced damage is due to prolonged electrical excitation of axons. Furthermore, since less than complete suppression of the induced neural activity by local anaesthetic engenders essentially complete sparing of all axons, our results suggest that the damage to individual axons derives, at least in part, from stimulation-induced global changes in the nerve.

Partial pressure of oxygen in brain and peripheral nerve during damaging electrical stimulation.

The studies were performed to elucidate the mechanism underlying the neural damage which may occur during prolonged electrical stimulation of either brain tissue or peripheral nerve. The partial pressure of oxygen (pO2) was measured in the sciatic nerve and the cerebral cortex of adult cats before and during direct, local electrical stimulation of these neural tissues, using stimulus parameters capable of inducing neural injury. pO2 was monitored by the polarographic method, employing a platinum microelectrode inserted into the tissue adjacent to or beneath the stimulating electrode. In the sciatic nerve there was no marked change in intrafascicular pO2 in three cats upon initiation of the electrical stimulation. In a fourth animal intraneural pO2 increased briefly upon initiation of the stimulation. In no case did the intrafascicular compartment of nerves become significantly hypoxic. In the cerebral cortex, the start of stimulation was accompanied by a significant increase (approximately 12-15 Torr) in intracortical pO2 beneath the stimulating electrode, and pO2 remained at or above the pre-stimulus value for the duration of the stimulation. These results show that extracellular hypoxia is unlikely to be a significant factor in the neural injury induced in brain or peripheral nerve by prolonged electrical stimulation.

Considerations for safety with chronically implanted nerve electrodes.

Electrical stimulation of cranial and peripheral nerves has been used to ameliorate a variety of neurologic disease states and neural injuries over the past 20 years. In this review, clinical applications and the histopathologic results of chronic implants in animals and humans are discussed, and the results of neural damage models developed at Huntington Medical Research Institutes are summarized. Chronically implanted electrode arrays may produce neural injury by either mechanical factors or by continuous, high-frequency electrical stimulation. The margin of safety to avoid electrically induced injury may be increased by minimizing the frequency or total stimulation time, and by the use of an intermittent duty cycle. The protocols presently being used for the stimulation of the vagus nerve to effect inhibition of seizures appear to have an adequate margin of safety.