Neuroprosthetics

The following sections are included:

• PREFACE

• CONTENTS

Excitable cells show a strongly nonlinear relationship between the transmembrane potential and the membrane current. In particular, after the membrane potential reaches threshold, the membrane potential follows a stereotyped wave shape called the action potential. Nevertheless, up to about 80% of the threshold level, the membrane potential and current can be described accurately using linear, or passive, models. In this chapter, we focus on the description of biological tissue as a target for electrical stimulation based on the passive properties of excitable cells. We first look at the relevant structure of the excitable cell with regard to electrical activation: the cell membrane, and the presence of ion channels. Then the resting potential is described and, finally, linear models for the response of excitable cells to intracellular and extracellular electrical stimuli are given for various kinds of cells.

This chapter provides an overview of the basic structure and physiology of the peripheral nerve and sense organs they innervate.

The goal of this chapter is to provide a general introduction to the central nervous system and serve as a framework for the reader to use in considering subsequent chapters that deal with particular components of this system. Given the vast complexity of the central nervous system, a single chapter falls very short of properly describing the anatomy and physiology of the various structures that comprise it. We have therefore chosen to describe only six structures in more detail: spinal cord, brainstem, cerebellum, thalamus, basal ganglia, and somatosensory and motor cortex. A brief description of the anatomy and physiology of each structure is provided. The discussion of each structure also includes a section related to neuroprosthetics. Throughout the chapter, we refer the reader to a number of other chapters in this book which describe the anatomy and physiology of additional structures within the central nervous system.

This chapter provides and overview of the structure of the sympathetic and parasympathetic nervous systems and examples of its function, particularly with respect to the cardiovascular system.

This chapter provides an overview of the basic anatomical, biochemical, physiological, and biomechanical properties of skeletal muscle and how these properties are affected by patterns of activity.

This chapter presents an overview of the functional anatomy and physiological substrates within the central nervous system that are involved in the planning and execution of voluntary motor activity. Many classic and seminal experiments that have lead to our present understanding of the mechanisms of motor control are described. This is done to support the experimental findings and to introduce some of the methods that are available to perform investigations in this field, including physiological psychology, electrophysiology, neuroanatomy, clinical pathology, and behavioral research. There are scores of excellent published research articles and books devoted to the topic of control of voluntary movement, so this chapter only provides an overview. The goal is to introduce some of the most relevant background information that a bioengineer or rehabilitation researcher would find useful to make progress in the field of motor neuroprosthetics.

The anatomy, physiology, and function of the healthy visual system are presented, as well as some of the more common eye diseases disrupting its normal function. Particular emphasis is placed on properties of the visual system that may be difficult to recreate through a neuroprosthesis, to provide the reader with a realistic sense of the challenges facing visual prosthesis designers.

The development and clinical success of auditory prostheses rests, in part, on our knowledge of the structure and function of the auditory system. This chapter reviews the mammalian auditory pathway, placing considerable emphasis on the anatomy and physiology of the cochlea. This background is necessary to understand the pathological implications of a sensorineural hearing loss and its amelioration via an auditory prosthesis.

In recent years our understanding of the organization of the cerebral cortex has changed. It is now known that the brain is capable of adaptation to environmental challenges (as in learning) and to functional disabilities produced by lesions. The existence of this process, often described as plasticity, stimulated the development of interventions geared to enhance plasticity when it plays a beneficial role and to inhibit it when it is detrimental. This chapter will review the use of transcranial magnetic stimulation, a noninvasive technique used to study and to modulate neuroplasticity in humans.

Victims of spinal cord injury may experience a recovery from or worsening of their initial post-trauma sensation and use of limb muscles and visceral organs. This chapter will review animal and human studies of spinal cord injury, describing the anatomical and physiological changes in the spinal cord that accompany the changes in somatic and autonomic functions in the post-trauma period. The locomotor and voiding functions will be addressed in detail due to their high impact on well-being of patients and availability of animal models to study the spinal mechanisms of these functions. The initial recovery of walking and bladder emptying abilities is often attributed to restored conduction at the site of spinal trauma. However, the long-term recovery is often very gradual and involves the on-going plasticity of spinal connections below the injury site. The recovery process can be accelerated by employing procedures of electrically-induced spinal training.

Electrical stimulation of peripheral nerve fibers is an important method to study physiological mechanisms and a technique to restore function to persons with neurological disorders. The processes underlying excitation of nerve fibers are presented through the development of quantitative models of the non-linear properties of the neuronal membrane and the response of nerve fibers to extracellular stimulation. Similarly, the properties of nerve fiber excitation are described and the origins of these properties are explained by reference to the quantitative models. The quantitative treatment of nerve fiber excitation provides the means for understanding the origins of experimentally observed phenomena and a set of tools that can be used to design stimulation paradigms that generate the desired pattern of excitation.

The main focus of this chapter is the bioelectrical, chemical, and physical theories underlying the recording of bioelectric potentials from the body, with emphasis placed on recording from the peripheral nervous system. Following a general overview of recording bioelectric signals, we start our discussion with a mathematical description of the extracellular electric field distribution produced by the active nerve fiber, and discuss implications from the model predicting how the electrode, its placement and recording configuration influences the recording. These models are based upon certain assumptions about the regular electrochemical properties of the system. We continue with a discussion of the properties of the electrolytic medium in which signals travel and make our way to the electrochemical interface in which they are picked up. Finally, we discuss the recording system that picks up the signal and methods to evaluate the quality of the signal and noise that exist in the recording.

The excitability of neurons is analyzed in examples for functional electrical stimulation of the cochlea, the retina, and the spinal cord.

A large number of biophysical models and quantitative methods are relevant to the design of Brain-Computer Interfaces and other neuroprosthetic devices that rely on recording neural activity in the CNS. These models describe different biological processes occurring across a range of scales. Here we focus on three that appear most relevant to the neural engineer. At the smallest scale we review models of extracellular potential spread around a spiking neuron as they relate to the sources of signal and noise encountered during microelectrode recordings. We overview a variety of signal detection and pattern recognition methods used in an attempt to reliably and automatically detect single unit activity from these recordings. Next, we look at the way the activity of very large populations of neurons aggregates to form the fields measured using far-field recordings from the skull or scalp. We conclude with a review of the representation of dynamic sensory and motor quantities in the activity of spiking neurons and small populations of neurons within the framework of encoding models and decoding methods.

Electrodes represent the direct interface between the biological tissue and the electronic system in neural prostheses. Within this chapter, the basic mechanisms at the electrode-electrolyte phase boundary are physico-chemically described and modeled in an equivalent circuit model with lumped parameters. Important measurement methods for electrode characterization are introduced. The most common recording and stimulation electrode materials are discussed with respect to their application. The chapter closes with an overview of novel techniques to optimize electrode properties with means of micro system technology.

Implantable devices for neuroprostheses should be as small as possible to minimize damage to target neural structures. Silicon integrated circuit and micromachining technology has enabled fabrication of very small, electrode arrays and complex electrical systems. However, application of this advanced technology requires development of new insulating materials for operation of theses electronic devices within the conductive, aqueous environment of animals and humans. These insulating biomaterials must be “biocompatible” (cause no foreign body response from the host) and “bioresistant” (remain unaffected by the host environment) from a few months up to 100 years for various applications. The results described are from “in progress” research under a long standing program funded by the NIH Neuroprosthesis Program. Commercially available and newly synthesized insulating biomaterials are being evaluated using a variety of sensitive measurements and accelerated degradation techniques. Wire insulators, ribbon cable interconnect insulators, silicon device surface encapsulants, encapsulants for alumina and quartz hybrid circuit substrates and assemblies, and encapsulants for silicon CMOS integrated circuit devices are being studied. Surface adhesion, material mechanical properties, and electrical performance, are being evaluated at 37°C, 90°C, and in-vivo. Silicones and fluoropolymers are the most stable materials yet studied.

Chemical Vapor Deposition (CVD) is a technique by which thin, conformal coatings can be synthesized and applied onto substrates with complex topographies and small overall dimensions. The process does not require any solvents and can be performed in a single step. Polymeric fluorocarbon and organosilicon thin films have been synthesized by two different CVD techniques, plasma-enhanced CVD and hot-filament CVD. Uniform, conformal films are produced on various substrates such as platinum wire, silicon microribbons, and neural probes. The synthesis of fluorocarbon-organosilicon copolymers, which offers the possibility of combining the useful properties of each material type in a single thin film, is also discussed.

Electrical stimulation is an effective albeit an unnatural, means of exciting neurons and axons, and certain precautions are necessary if tissue injury is to be avoided. Prolonged electrical stimulation can inflict neural injury that is clearly evident in histologic sections, and also may induce more subtle (and usually reversible) changes, included prolonged depression of electrical excitability. This chapter reviews the electrochemistry of charge injection into tissue by metal stimulating electrodes and discusses several mechanisms by which the stimulation might induce neuronal injury, both in the central nervous system and in peripheral nerves.

Chronic stimulation of a skeletal muscle over a period of weeks induces biochemical, physiological, and structural changes that enable it to sustain new and demanding working conditions without fatigue. The discovery of this adaptive response has opened up a number of clinical applications in which muscles are activated artificially or redeployed as a functional biomaterial. The issues are explored here in relation to the use of such muscles to provide cardiac assistance in patients with heart failure. An appropriate and informed choice of stimulation pattern is crucially important to establish and to maintain the long-term viability, dynamic characteristics, working capacity, and resistance to fatigue needed in this application. Furthermore, mobilization of the muscle as a functional graft involves a partial loss of vascular supply that renders it vulnerable to ischaemic damage when it is subjected to the metabolic challenge of stimulation. Such damage, and the resultant deterioration in function, can be avoided by the use of true vascular delay and/or prestimulation protocols. In recent years advances in the associated basic science have provided a far better understanding of the problems involved in cardiac assistance from skeletal muscle, and corresponding refinement of the operative procedures should enable the full potential of these techniques to be realized.

In this chapter we focus on technology to activate peripheral motor nerves electrically. Two important concepts are stressed: 1) the closer the electrode is to the target tissue the easier it is to isolate the applied electric field to a smaller region, and 2) the effect of the applied electric field is, generally speaking, always the greatest on the largest myelinated axons experiencing the applied electric field. These concepts are applicable to other neural systems. Motor nerves can be activated through electrodes placed on the surface of the skin, on the surface of the muscle, in the muscle, on the motor nerve, or in the motor nerve. All electrodes must satisfy the requirements of material compatibility, mechanical compatibility and the ability to transfer the required electrical charges without tissue or material deterioration. The choice of electrode materials and geometric design are determined by these factors and by the intended location on the nerve or muscle. Specific designs, tissue reactions, and applications are described herein. Electrodes placed on muscles produce single muscle activation. Nerve electrodes can have the advantage of activating multiple muscles. Selective stimulation of peripheral nerve fibers for effecting specific muscle activation or specific motor function is discussed in the section on nerve electrodes.

Functional neuromuscular stimulation (FNS) systems for neuroprostheses often involve implanted peripheral nerve interfaces. The utility of these implanted electrode systems includes stimulation to evoke muscle contractions, but also extends to recording from sensory afferents to provide sensory information for feedback control of the FNS systems. Peripheral nerve recordings can also provide cognitive feedback to the user of the FNS system and obtain user generated volitional commands. This chapter describes these sensory applications and provides a summary of the current state of the art in this field. The narrative begins with an explanation of the concepts and rationale behind recording peripheral nerve signals for FNS systems, followed by a survey of current electrode styles, including descriptions and illustrations of many of the devices that researchers have fabricated. This is followed by a presentation of the current applications of sensory recording for use with various FNS systems such as apnea monitoring, correction of drop foot, and restoration of grasp function. The chapter concludes with a statement of our opinion concerning the trends and future directions in the field.

While stimulating electrode charge delivery and safety performance depends to a large extent on electrochemistry, the utility of neural stimulation devices is geometry dependent at several scales from nanometer to millimeter. The molecular scale of the chemical reaction, the tortuosity properties of the surface, micron scale micromachining, through to the gross physical features of an electrode system all contribute to the distribution and dynamics of current surrounding the device. Here, we explore the effects of geometry at some of these scales with examples of some of the more common forms that electrodes take.

Recording electrodes provide the direct brain interface for an emerging class of neuroprosthetic systems. This chapter focuses on implantable microelectrode arrays for recording neuronal action potentials and local field potentials. The fundamental operating principles of these types of microelectrodes are discussed within the broader context of neural implant systems. Microelectrode performance is discussed in relation to electrode impedance, electrode size and shape, and electrode materials. Overall considerations of neural implant systems for long-term recording in humans are presented. The chapter concludes by presenting a brief survey of different types of microelectrodes.

In this chapter we review the development of spinal cord stimulation and recording techniques in the context of neuroprosthetics in general. The decade 1965-1975 saw rapid and adventurous developments in implantable stimulators. Many of the present-day systems are based on this early work. In the intervening 25 years over 60,000 neuroprostheses have been implanted, 50,000 of them being cochlear stimulators. Dorsal column stimulators, phrenic pacers, deep-brain stimulators and sacral root stimulators each contributed about 2,000-3,000 to the total. Intraspinal stimulation was first tried in 1970 with promising results, however interest turned to peripheral nerve stimulators. In the intervening time, the advantages and disadvantages of peripheral stimulation have been established. There is renewed interest in intraspinal stimulation because it is recognized that the activation of “command centers” may allow coordinated, synergistic outputs to be elicited, which could potentially get around some of the limitations of peripheral stimulation. Sensory signals recorded from spinal roots may be useful in controlling stimulation but stability remains a difficult issue.

The review starts with the conceptual description of control systems needed for efficient motor neuroprostheses; this is with a scheme of a hybrid, hierarchical controller. The review comprises an overview of control systems for movement, modeling of the skeletal and musculotendonal systems, and identification techniques needed for determining the parameters. It also presents elements of nonanalytical techniques and rule-based methodologies used for neuroprosthesis. The final part describes the actual status of available motor neuroprostheses.

Motor neuroprostheses can provide standing, stepping, and walking functions for individuals with hemiplegia or paraplegia; and can provide grasping and reaching functions for individuals with tetraplegia. Motor neuroprostheses are designed to give the user control of the patterned electrical activation of his or her muscles. There are over 30 different clinically-deployed upper or lower extremity neuroprostheses reported in the literature, ranging in complexity from single channel surface stimulators to multi-channel implanted stimulator-telemeter systems. Significant functional gains have been reported by these applications and there are a number of commercially available systems. Motor neuroprostheses provide a tool for improving the independence of disabled individuals in a manner that cannot be achieved through any other means.

This chapter briefly reviews the pathological response of the cochlea following deafness, and then presents a detailed overview of the design of a contemporary cochlear implant.

Neuromodulation and neurostimulation have evolved from experimental techniques to clinical practice. The majority of implantations are still performed in specialised units and centers but an increasing number of implants are performed worldwide with excellent success. However both techniques using electrical current as the driving force make use of different mechanisms and, hence, address different categories of patients. Electrical stimulation of the anterior roots of the sacral nerves is used to provoke bladder contraction as a direct effect of nerve stimulation. High amplitudes are necessary and therefore the technique is limited to patients with a spinal cord injury. Neuromodulation uses a modulatory effect of electrical current on the controlling reflexes of the lower urinary tract inflow into the central nervous system. Therefore this technique can be used in conditions seemingly as different as overactive bladder or non-contracting bladder. The excellent clinical results with long-term follow-up makes these techniques a valuable contribution in the field of neuro-urology.

Deep brain stimulation (DBS), which utilizes chronically implanted electrodes in the central nervous system, has proven effective for a wide range of neurological conditions. Despite its effectiveness however, little is known regarding its mechanisms of action. Theoretically, the precise delivery of electrical stimulation is capable of achieving spatial and temporal resolution that current pharmacological treatments cannot begin to approach. Currently however, DBS therapies are relatively crude and targets for DBS are limited to those structures that had previously been the target of destructive surgeries. Increasingly sophisticated electronic devices and electrodes, better understanding of the effects of DBS on neurons and systems of neurons, and increased knowledge of pathophysiological mechanisms should lead to better and more far ranging neurological treatments. Finally, a better understanding of how DBS mitigates symptoms in treated patients may provide further insight into the physiology and pathophysiology of the central nervous system.

The development of silicon substrate recording arrays for the study of several neurons simultaneously has given neuroscientists and prosthesis engineers new options of site distribution and overall geometric shape. The analysis methods for utilizing many channels of data recorded from independent channels has been developed extensively. New silicon electrodes with close-spaced sites present data that has considerable overlap of both neural channels and noise. This characteristic calls upon mathematical methods long utilized for analysis of signals from electromagnetic antenna arrays. Following this line of analysis for tetrode type electrodes has provided benefits for cell signal separation but utilization with larger and more precisely distributed sites builds on these benefits by providing methods to extract spatial cell distributions, improved signal-to-noise ratios for cell signals and the ability to extract more broadly correlated signals such as evoked potentials.

Spinal cord injuries often cause paralysis of the respiratory muscles. Loss of inspiratory muscle function, including the diaphragm, results in respiratory failure. Loss of expiratory muscle function markedly reduces cough effectiveness. Electrical stimulation techniques have been developed to pace the diaphragm and restore inspiratory muscle function. Bilateral phrenic nerve pacing is commercially available and a clinically accepted technique to restore respiratory muscle function, freeing patients from dependence upon mechanical ventilation. Clinical trials of laparoscopically placed intramuscular diaphragm electrodes are underway as an alternative to phrenic nerve pacing. This method would alleviate the need for a thoracotomy and reduce the possibility of phrenic nerve injury. Clinical studies are also in progress to restore inspiratory intercostal muscle function. Combined intercostal and unilateral diaphragm pacing holds promise as an alternative method of ventilatory support in patients with only unilateral diaphragm function and therefore not candidates for phrenic nerve pacing. Finally, a number of methods of expiratory muscle activation are being evaluated to restore expiratory muscle function. Restoration of an effective cough mechanism may reduce the incidence of respiratory tract infection and atelectasis, common causes of morbidity and mortality in the spinal cord injury population.

Basic research in areas related to motor neuroprostheses has led to a large number of technical and theoretical advancements over the past 30 years, some of which have migrated into clinically deployed systems but many of which have not. This chapter discusses recent advances in a number of different areas (activation, command and control algorithms, and devices) to assess the likelihood of their eventual adoption in deployed systems. Opportunities for expanding the functional benefits of neuroprostheses to additional groups of people with neurological disabilities are evaluated and the future role of neuroprostheses relative to other approaches (such as CNS regeneration) for functional restoration is considered. Widespread clinical deployment of neuroprostheses will likely require significant industrial participation, so barriers and opportunities in the commercial sector are discussed.

Short of limb regeneration, an ideal prosthesis for amputees is one that is interfaced to peripheral nerve stumps. In this chapter, we discuss the feasibility of interfacing an artificial arm to human nerve stumps. In principle, the concept is simple and straightforward. In practice however, changes in the central and peripheral nervous system postaxotomy/amputation, the need for an interface that shows both electrochemical and mechanical biocompatibility, the potential for nerve injury, the need to record nerve impulses and stimulate sensory neurons over a period of years all add to our difficulties. These issues will be addressed together with a brief discussion on the progress to date with experiments involving human amputee subjects.

The task of restoring standing and stepping in paraplegia using conventional functional electrical stimulation techniques has been very challenging. Both functions require the coordinated activation of several muscles, thereby necessitating the implantation of multiple intramuscular or epinerual electrodes throughout the limbs and the subcutaneous tunneling of long leads towards a common stimulator. This chapter discusses the development and potential future use of two novel electrical stimulation systems for restoring motor function to the lower extremities; namely, leadless injectable intramuscular stimulators (BIONs) and intraspinal microstimulation (ISMS). The current status of each system is presented with special emphasis given to ISMS.

Four emerging applications of electrical stimulation to treat bladder dysfunction are described. The first application concerns selective stimulation of the small diameter nerve fibers in the sacral nerve roots. This allows for near physiological bladder emptying in spinal cord injured patients. The second application treats detrusor overactivity by stimulating the pudendal nerve using an injectable stimulator (BION). Treatment of neurogenic detrusor overactivity is the aim of the third application. A closed loop stimulation system, consisting of a sensor and stimulator, is used and allows for conditional stimulation. Finally an application to treat mixed incontinence is described.

Blindness is a severe handicap and a huge socioeconomic burden. Several groups around the world are evaluating the feasibility of creating visual perception in blind individuals by electrical stimulation of neuronal tissue along the visual pathways. Stimulation locations include the retina, the optic nerve and the visual cortex. These preliminary experiments are meant to study the safety and efficacy of visual prostheses. Such prostheses will convert images into electrical pulses and then use these pulses to stimulate the neurons in a pattern representative of the image. Controlled electrical pulses applied to neurons through each electrode site would produce the perception of a dot of light corresponding to the active electrode. Using electrical stimulation through patterns of electrodes would allow the patient to see many dots of light and by piecing these dots together, see an image; similar to how dot-matrix printers use small ink dots on paper to form coherent images. The intent of this chapter is to provide an overview of past, current, and future concepts of visual prostheses. It will begin with a description of each of the cortical, retinal, and optic nerve designs. Attempts at implanting electronic devices at various parts along the visual pathways will be discussed. Both major achievements and obstacles remaining will be summarized. Then, the chapter will concentrate on general considerations that are related to all or most of visual prostheses. Given that intact neurons along the visual pathways can be found in almost all blind patients, only our lack of experience and capabilities in physiology, biocompatibility and device-tissue interfacing is preventing us from stimulating them in a safe and effective manner. We believe that as our knowledge about how to stimulate neurons with microelectronics increases and as microelectronics and material sciences continue to evolve, we should one day be able to restore vision to the blind. Given the advances that have been made in this field we can only hope that the day such devices are widely used is in the near future and not decades away.

This chapter reviews historical and current techniques designed to provide auditory cues to the severe-profoundly deaf via neural prostheses designed to stimulate sites within the central auditory pathway.

Disorders of the peripheral vestibular system are relatively common and often result in severely impaired mobility, blurred vision and debilitating attacks of vertigo. Preliminary research in the area of vestibular neuroprosthetics provides hope that some vestibular impaired patients may eventually be helped by providing alternative motion cues or, ultimately, restoring motion-sensitive vestibular afferent inputs to the brain. Conceptually, vestibular prosthetics are relatively simple, consisting of 4 principal elements: a power source, sensors, a signal processor, and stimulators. These 4 elements can be combined in various different ways to provide information to the nervous system. This chapter reviews the current state of vestibular sensory substitution devices and peripheral vestibular neuroprosthetics. While preliminary work indicates clinical potential, numerous unresolved technical challenges remain to be addressed.

A Brain-Computer-Interface (BCI) based on learning of self-regulation of slow cortical potentials (SCP) is described. This Thought-Translation-Device (TTD) allows completely paralyzed and locked-in patients to select letters or words with their SCPs recorded non-invasively at the scalp. The neurophysiological basis of SCPs and the hierarchically ordered training steps for locked-in patients were tested experimentally on 7 paralyzed and artificially respirated patients with end-stage amyotrophic lateral sclerosis.

Neuromotor prostheses are a type of brain-machine interface (BMI) that seek to extract signals from the central or peripheral nervous system and deliver them to control devices. A brain-machine interface is necessary to detect activity that can be voluntarily modulated for use as a motor control signal. It is generally accepted that electrical potentials are the most valuable sources of information. Neural commands for voluntary movement are essentially issued as electrical signals produced by the spiking (action potentials) and synaptic input of individual neurons; both can be recorded with varying degrees of fidelity and difficulty. The goal is to be able to detect signals that have the largest amount of information about movement and that change about as rapidly as movement commands themselves change. Clearly, recording at the source of the motor commands most readily fulfills these requirements, but indirect recordings of surrogate signals can provide an alternative or supplemental source, if one can learn to make indirect signals mimic motor commands. The decoding methods for use in neuromotor prostheses are the culmination of many years of basic research on the motor system. Whereas recovering movement dynamics and kinematics from neural activity alone comprises a feat of basic science, their use as a control signal marks a shift to applied neuroprosthetics. In this chapter we review mathematical algorithms that have been tested in prototypes of intracortical neuromotor prostheses. ‘Closed-loop’ refers to the situation wherein the subject is provided access to recovered movement information, and is required to use this prediction signal in a behaviorally useful manner. This access may be afforded visually (neurally derived cursor trajectories), mechanically (as in stimulation of muscles via implanted electrodes), or any number of output devices. We will consider several features that are unique to the closed-loop context of online control, including those specific to use in paralyzed human patients. We consider here the advantages and disadvantages of field potentials and spikes; in the final section of the chapter, we argue that a principled combination of all available information channels, processed by a multiplicity of decoding algorithms, will result in the most effective neuromotor prosthesis.

Much of our understanding of brain function and organization has been provided via the single microelectrode. Single microelectrode studies of sensory and motor processing in the brain have also inspired the development of neuroprosthetic approaches that someday may aid the deaf, blind, or paralyzed. Necessarily complex, such approaches will require sophisticated multi-electrode arrays implanted into one or more regions of the brain. Today, devices available for implantation into the brain have multiple electrode sites, are chronically implantable, and can include circuitry for on-board signal processing. The utility of multi-electrode arrays to either electrically stimulate or record within the brain has been well-demonstrated in basic research animal studies. However, such devices are not yet regularly incorporated in the clinical arena. Although the complex neural interfaces available today are impressive, they have yet to demonstrate the necessary longevity required or to meet the other stringent requirements for human implantation. Only when electrode systems can be made to function reliably and consistently for the lifetime of an implanted subject will these gains be possible. In this chapter, a brief review of known and postulated failure modes and device limitations are presented in an effort to define a working plan for the development of more effective devices in the future. This plan incorporates the best of currently available neural engineering technologies with new and as of yet undeveloped advances. By understanding the failure modes of previous systems and anticipating the requirements of advanced systems, a plan for development of the next generation of cortical implants can be outlined.

A brief treatise of biocompatibility concepts, test procedures and their interpretation, including material selection, characterization, design and sterilization, with a foray into the history, rules and regulations one must understand and follow to register a neuroprosthetic medical device before commercialization is presented. An overview of the process that guides selection of the regulatory route is summarized with the basics of the investigational device exemptions, pre-market notification, and pre-market approval process included.

The following sections are included:

• INDEX