Robinson, C.J. "Rehabilitation Engineering, Science, and Technology The Electrical Engineering Handbook Ed. Richard C. Dorf Boca raton crc Press llc. 2000
Robinson, C.J. “Rehabilitation Engineering, Science, and Technology” The Electrical Engineering Handbook Ed. Richard C. Dorf Boca Raton: CRC Press LLC, 2000
Rehabilitation Engineering, Science and Technology 119.1 Rehabilitation Concepts 119.2 Engineering Concepts in Sensory Rehabilitation 119.3 Engineering Concepts in Motor Rehabilitation Charles J. Robinson 119.4 Engineering Concepts in Communications Disorders Louisiana tech university 119.5 Appropriate Technology Overton Brooks VA Medical Center 119.6 The Future of Engineering in Rehabilitation Rehabilitation engineering requires a multidisciplinary effort. To put rehabilitation engineering into its proper context, we need to review some of the other disciplines with which rehabilitation engineers must be familiar. Robinson [1993] has reviewed or put forth the following working definitions and discussions ehabilitation: The(Re)integration of an individual with a disability into society. This can be done either by enhancing existing capabilities or by providing alternative means to perform various functions or Rehabilitation engineering: The application of science and technology to ameliorate the handicaps of indi- als with disabilities [Reswick, 1982]. In actual practice, many individuals who say that they practice rehabilitation engineering are not engineers by training. While this leads to controversies from practi- tioners with traditional engineering degrees, it also has the de facto benefit of greatly widening the scope of what is encompassed by the term"rehabilitation engineering Rehabilitation medicine: A clinical practice that focuses on the physical aspects of functional recovery, but that also considers medical, neurological and psychological factors. Physical therapy, occupational ther apy, and rehabilitation counseling are professions in their own right On the sensory-motor side, other medical and therapeutical specialties practice rehabilitation in vision, audition, and speech Rehabilitation technology (or Assistive technology): Narrowly defined, the selection, design, or manufactur of augmentative or assistive devices that are appropriate for the individual with a disability. Such devices are selected based on the specific disability, the function to be augmented or restored, the user's wishes the clinicians preferences, cost, and the environment in which the device will be used. Rehabilitation science: The development of a body of knowledge, gleaned from rigorous basic and clinical research, that describes how a disability alters specific physiological functions or anatomical structures, and that details the underlying principles by which residual function or capacity can be measured and used to restore function of individuals with disabilities c 2000 by CRC Press LLC
© 2000 by CRC Press LLC 119 Rehabilitation Engineering, Science, and Technology 119.1 Rehabilitation Concepts 119.2 Engineering Concepts in Sensory Rehabilitation 119.3 Engineering Concepts in Motor Rehabilitation 119.4 Engineering Concepts in Communications Disorders 119.5 Appropriate Technology 119.6 The Future of Engineering in Rehabilitation Rehabilitation engineering requires a multidisciplinary effort. To put rehabilitation engineering into its proper context, we need to review some of the other disciplines with which rehabilitation engineers must be familiar. Robinson [1993] has reviewed or put forth the following working definitions and discussions. Rehabilitation: The (Re)integration of an individual with a disability into society. This can be done either by enhancing existing capabilities or by providing alternative means to perform various functions or to substitute for specific sensations. Rehabilitation engineering: The application of science and technology to ameliorate the handicaps of individuals with disabilities [Reswick, 1982]. In actual practice, many individuals who say that they practice rehabilitation engineering are not engineers by training. While this leads to controversies from practitioners with traditional engineering degrees, it also has the de facto benefit of greatly widening the scope of what is encompassed by the term “rehabilitation engineering.” Rehabilitation medicine: A clinical practice that focuses on the physical aspects of functional recovery, but that also considers medical, neurological and psychological factors. Physical therapy, occupational therapy, and rehabilitation counseling are professions in their own right. On the sensory-motor side, other medical and therapeutical specialties practice rehabilitation in vision, audition, and speech. Rehabilitation technology (or Assistive technology): Narrowly defined, the selection, design, or manufacture of augmentative or assistive devices that are appropriate for the individual with a disability. Such devices are selected based on the specific disability, the function to be augmented or restored, the user’s wishes, the clinician’s preferences, cost, and the environment in which the device will be used. Rehabilitation science: The development of a body of knowledge, gleaned from rigorous basic and clinical research, that describes how a disability alters specific physiological functions or anatomical structures, and that details the underlying principles by which residual function or capacity can be measured and used to restore function of individuals with disabilities. Charles J. Robinson Louisiana Tech University Overton Brooks VA Medical Center
119.1 Rehabilitation Concepts Effective rehabilitation engineers must be well versed in all of the areas described above because they generally work in a team setting, in collaboration with physical and occupational therapists, orthopedic surgeons, physical medicine ecialists, and/or neurologists. Some rehabilitation engineers are interested in certain activities that we do in the course of a normal day that could be summarized as activities of daily living(ADL). These include eating, toileting, ombing hair, brushing teeth, reading, etc. Other engineers focus on mobility and the limitations to mobility. Mobility can be personal (e.g, within a home or office)or public(automobile, public transportation, accessibility questions in buildings). Mobility also includes the ability to move functionally through the environment. Thus, the question of mobility is not limited to that of getting from place to place, but also includes such questions as whether one can reach an object in a particular setting or whether a paralyzed urinary bladder can be made functional again Barriers that limit mobility are also studied. For example, an ill-fitted wheelchair cushion or support system will nost assuredly limit mobility by reducing the time that an individual can spend in a wheelchair before he or she must vacate it to avoid serious and difficult-to-heal pressure sores. Other groups of rehabilitation engineers deal with sensory disabilities, such as sight or hearing, or with communications disorders, both on the production side g, the nonvocal)or on the comprehension side. For any given client, a rehabilitation engineer might have all of these concerns to consider (ie, ADLs, mobility, sensory, and communication dysfunctions A key concept in physical or sensory rehabilitation is that of residual function or residual capacity. Such a concept implies that the function or sense can be quantified, that the performance range of that function or sense is known in a nonimpaired population, and that the use of residual capacity by a disabled individual should be encouraged. These measures of human performance can be made subjectively by clinicians or objectively by some rather clever computerized test devices. A rehabilitation engineer asks three key questions: Can a diminished function or sense be successfully augmented? Is there a substitute way to return the function or to restore a sense? And is the solution appropriate and cost-effective? These questions give rise to two important rehabilitation concepts: orthotics and prosthetics. An orthosisis an appliance that aids an existing function. a prosthesisprovides a substitute. An artificial limb is a prosthesis, as is a wheelchair. An ankle brace is an orthosis so are eyeglasses. In fact, eyeglasses might well be the penultimate rehabilitation device. They are inexpensive, have little social stigma, and are almost completely unobtrusive to the user. They have let many millions of individuals with correctable vision problems lead productive lives. But in essence, a pair of eyeglasses is an optical device, governed by traditional equations of physical optics. Eyeglasses can be made out of simple glass(from a raw material as bundant as the sands of the earth) or complex plastics such as those that are ultraviolet sensitive. They can be ground by hand or by sophisticated computer-controlled optical grinders. Thus, crude technology can restore functional vision. Increasing the technical content of the eyeglasses(either by material or manufacturing method) in most cases will not increase the amount of function restored, but it might make the glasses cheaper, lighter, and more prone to be used 119.2 Engineering Concepts in Sensory Rehabilitation Of the five traditional senses, vision and hearing best define the interactions that permit us to be human. These two senses are the main input channels through which data with high information content can flow. We read we listen to speech or music; we view art. A loss of one or the other of these senses (or both) can have a devastating impact on the individual affected Rehabilitation engineers attempt to restore the functions of these senses,either through augmentation or via sensory substitution systems. Eyeglasses and hearing aids are examples of augmentative devices that can be used if some residual capacity remains. A major area of rehabil itation engineering research deals with sensory substitution systems [ Kaczmarek et al, 1991] The visual system has the capability to detect a single photon of light, yet also has a dynamic range that can respond to intensities many orders of magnitude greater. It can work with high contrast items and with those of almost no contrast, and across the visible spectrum of colors. Millions of parallel data channels form the optic nerve that comes from an eye; each channel transmits an asynchronous and quasi-random(in time stream of binary pulses. While the temporal coding on any one of these channels is not fast(on the order c 2000 by CRC Press LLC
© 2000 by CRC Press LLC 119.1 Rehabilitation Concepts Effective rehabilitation engineers must be well versed in all of the areas described above because they generally work in a team setting, in collaboration with physical and occupational therapists, orthopedic surgeons, physical medicine specialists, and/or neurologists. Some rehabilitation engineers are interested in certain activities that we do in the course of a normal day that could be summarized as activities of daily living (ADL). These include eating, toileting, combing hair, brushing teeth, reading, etc. Other engineers focus on mobility and the limitations to mobility. Mobility can be personal (e.g., within a home or office) or public (automobile, public transportation, accessibility questions in buildings). Mobility also includes the ability to move functionally through the environment. Thus, the question of mobility is not limited to that of getting from place to place, but also includes such questions as whether one can reach an object in a particular setting or whether a paralyzed urinary bladder can be made functional again. Barriers that limit mobility are also studied. For example, an ill-fitted wheelchair cushion or support system will most assuredly limit mobility by reducing the time that an individual can spend in a wheelchair before he or she must vacate it to avoid serious and difficult-to-heal pressure sores. Other groups of rehabilitation engineers deal with sensory disabilities, such as sight or hearing, or with communications disorders, both on the production side (e.g., the nonvocal) or on the comprehension side. For any given client, a rehabilitation engineer might have all of these concerns to consider (i.e., ADLs, mobility, sensory, and communication dysfunctions). A key concept in physical or sensory rehabilitation is that of residual function or residual capacity. Such a concept implies that the function or sense can be quantified, that the performance range of that function or sense is known in a nonimpaired population, and that the use of residual capacity by a disabled individual should be encouraged. These measures of human performance can be made subjectively by clinicians or objectively by some rather clever computerized test devices. A rehabilitation engineer asks three key questions: Can a diminished function or sense be successfully augmented? Is there a substitute way to return the function or to restore a sense? And is the solution appropriate and cost-effective? These questions give rise to two important rehabilitation concepts: orthotics and prosthetics. An orthosis is an appliance that aids an existing function. A prosthesis provides a substitute. An artificial limb is a prosthesis, as is a wheelchair. An ankle brace is an orthosis; so are eyeglasses. In fact, eyeglasses might well be the penultimate rehabilitation device. They are inexpensive, have little social stigma, and are almost completely unobtrusive to the user. They have let many millions of individuals with correctable vision problems lead productive lives. But in essence, a pair of eyeglasses is an optical device, governed by traditional equations of physical optics. Eyeglasses can be made out of simple glass (from a raw material as abundant as the sands of the earth) or complex plastics such as those that are ultraviolet sensitive. They can be ground by hand or by sophisticated computer-controlled optical grinders. Thus, crude technology can restore functional vision. Increasing the technical content of the eyeglasses (either by material or manufacturing method) in most cases will not increase the amount of function restored, but it might make the glasses cheaper, lighter, and more prone to be used. 119.2 Engineering Concepts in Sensory Rehabilitation Of the five traditional senses, vision and hearing best define the interactions that permit us to be human. These two senses are the main input channels through which data with high information content can flow. We read; we listen to speech or music; we view art. A loss of one or the other of these senses (or both) can have a devastating impact on the individual affected. Rehabilitation engineers attempt to restore the functions of these senses, either through augmentation or via sensory substitution systems. Eyeglasses and hearing aids are examples of augmentative devices that can be used if some residual capacity remains. A major area of rehabilitation engineering research deals with sensory substitution systems [Kaczmarek et al., 1991]. The visual system has the capability to detect a single photon of light, yet also has a dynamic range that can respond to intensities many orders of magnitude greater. It can work with high contrast items and with those of almost no contrast, and across the visible spectrum of colors. Millions of parallel data channels form the optic nerve that comes from an eye; each channel transmits an asynchronous and quasi-random (in time) stream of binary pulses. While the temporal coding on any one of these channels is not fast (on the order of
200 bits per second or less), the capacity of the human brain to parallel process the entire image is faster than any supercomputer yet built. If sight is lost, how can it be replaced? A simple pair of eyeglasses will not work, because either the sensor he retina), the communication channel(the optic nerve and all of its relays to the brain), or one or more essential central processors(the occipital part of the cerebral cortex for initial processing: the parietal and other cortical areas for information extraction) has been damaged. For replacement within the system, one must determine where the visual system has failed and whether a stage of the system can be artificially bypassed. If one uses another sensory modality(e.g, touch or hearing) as an alternate input channel, one must determine whether there is sufficient bandwidth in that channel and whether the higher-order processing hierarchy is plastic enough to process information coming via a different route. While the above discussion might seem just philosophical, it is more than that. We normally read printed text with our eyes. We recognize words from their(visual) letter combinations We comprehend what we read via a mysterious processing in the parietal and temporal parts of the cerebral cortex. Could we perhaps read and comprehend this text or other forms of writing through our fingertips with an appropriate interface? The nswer,surprisingly, is yes! And, the adaptation actually goes back to one of the earliest applications of coding hat of the development of Braille. braille condenses all 3 dots(26 combinations), with certain combinations reserved as indicators for the next character(such as a lumber indicator) or for special contractions. Trained readers of Braille can read over 250 words per minute of grade 2 Braille(as fast as most sighted readers can read printed text). Thus, the Braille code is in essence a rehabilitation engineering concept where an alternate sensory channel is used as a substitute and where a ng scheme has been employed. Rehabilitation engineers and their colleagues have designed other ways to read text. To replace the retina as a sensor element, a modern high-resolution, high-sensitivity, fast-imaging sensor(CCD, etc. )is employed to capture a visual image of the text. One method, used by various page scanning devices, converts the scanned image to text by using optical character recognition schemes, and then outputs the text as speech via text-to- speech algorithms. This machine essentially recites the text, much as a sighted helper might do when reading aloud to the blind individual. The user of the device is thus freed of the absolute need for a helper. Such independence is often the goal of rehabilitation Perhaps the most interesting method presents an image of the scanned data directly to the visual cortex or retina via an array of implantable electrodes that are used to electrically activate nearby cortical or retin structures. The visual cortex and retina are laid out in topographic fashion such that there is an orderly mapping of the signal from different parts of the visual field to the retina, and from the retina to corresponding parts of the occipital cortex. The goal of stimulation is to mimic the neural activity that would have been evoked had the signal come through normal channels. And, such stimulation does produce the sensation of light. Since the"image"stays within the visual system, the rehabilitation solution is said to be modality specificHowever, substantial problems dealing with biocompatibility and image processing and reduction remain in the design of the electrode arrays and processors that serve to interface the electronics and neurological tissue. Deafness is another manifestation of a loss of a communication channel, this time for the sense of hearing Totally deaf individuals use vision as a substitute input channel when communicating via sign language(also a substitute code), and can sign at information rates that match or exceed that of verbal communication. Hearing aids are now commercially available that can adaptively filter out background noise(a predictable signal)while amplifying speech(unpredictable)using autoregressive, moving average(ARMA) signal process 8. With the recent advent of powerful digital signal processing chips, true digital hearing aids are now available Previous analog aids, or digitally programmable analog aids, provided a set of tunable filters and amplifiers to cover the low-, mid-, and high-frequency ranges of the hearing spectrum. But the digital aids can be specifically and easily tailored (i. e, programmed) to compensate for the specific losses of each individual client across the frequency continuum of hearing, and still provide automatic gain control and one or more user-selectable settings that have been adjusted to perform optimally in differing noise environments An exciting development is occurring outside the field of rehabilitation that will have a profound impact on the ability of the deaf to comprehend speech. Electronics companies are now beginning to market universal translation aids for travelers, where a phrase spoken in one language is captured, parsed, translated, and restated c 2000 by CRC Press LLC
© 2000 by CRC Press LLC 200 bits per second or less), the capacity of the human brain to parallel process the entire image is faster than any supercomputer yet built. If sight is lost, how can it be replaced? A simple pair of eyeglasses will not work, because either the sensor (the retina), the communication channel (the optic nerve and all of its relays to the brain), or one or more essential central processors (the occipital part of the cerebral cortex for initial processing; the parietal and other cortical areas for information extraction) has been damaged. For replacement within the system, one must determine where the visual system has failed and whether a stage of the system can be artificially bypassed. If one uses another sensory modality (e.g., touch or hearing) as an alternate input channel, one must determine whether there is sufficient bandwidth in that channel and whether the higher-order processing hierarchy is plastic enough to process information coming via a different route. While the above discussion might seem just philosophical, it is more than that. We normally read printed text with our eyes. We recognize words from their (visual) letter combinations. We comprehend what we read via a mysterious processing in the parietal and temporal parts of the cerebral cortex. Could we perhaps read and comprehend this text or other forms of writing through our fingertips with an appropriate interface? The answer, surprisingly, is yes! And, the adaptation actually goes back to one of the earliest applications of coding theory — that of the development of Braille. Braille condenses all text characters to a raised matrix of 2 by 3 dots (26 combinations), with certain combinations reserved as indicators for the next character (such as a number indicator) or for special contractions. Trained readers of Braille can read over 250 words per minute of grade 2 Braille (as fast as most sighted readers can read printed text). Thus, the Braille code is in essence a rehabilitation engineering concept where an alternate sensory channel is used as a substitute and where a recoding scheme has been employed. Rehabilitation engineers and their colleagues have designed other ways to read text. To replace the retina as a sensor element, a modern high-resolution, high-sensitivity, fast-imaging sensor (CCD, etc.) is employed to capture a visual image of the text. One method, used by various page scanning devices, converts the scanned image to text by using optical character recognition schemes, and then outputs the text as speech via text-tospeech algorithms. This machine essentially recites the text, much as a sighted helper might do when reading aloud to the blind individual. The user of the device is thus freed of the absolute need for a helper. Such independence is often the goal of rehabilitation. Perhaps the most interesting method presents an image of the scanned data directly to the visual cortex or retina via an array of implantable electrodes that are used to electrically activate nearby cortical or retinal structures. The visual cortex and retina are laid out in topographic fashion such that there is an orderly mapping of the signal from different parts of the visual field to the retina, and from the retina to corresponding parts of the occipital cortex. The goal of stimulation is to mimic the neural activity that would have been evoked had the signal come through normal channels. And, such stimulation does produce the sensation of light. Since the “image” stays within the visual system, the rehabilitation solution is said to be modality specific. However, substantial problems dealing with biocompatibility and image processing and reduction remain in the design of the electrode arrays and processors that serve to interface the electronics and neurological tissue. Deafness is another manifestation of a loss of a communication channel, this time for the sense of hearing. Totally deaf individuals use vision as a substitute input channel when communicating via sign language (also a substitute code), and can sign at information rates that match or exceed that of verbal communication. Hearing aids are now commercially available that can adaptively filter out background noise (a predictable signal) while amplifying speech (unpredictable) using autoregressive, moving average (ARMA) signal processing. With the recent advent of powerful digital signal processing chips, true digital hearing aids are now available. Previous analog aids, or digitally programmable analog aids, provided a set of tunable filters and amplifiers to cover the low-, mid-, and high-frequency ranges of the hearing spectrum. But the digital aids can be specifically and easily tailored (i.e., programmed) to compensate for the specific losses of each individual client across the frequency continuum of hearing, and still provide automatic gain control and one or more user-selectable settings that have been adjusted to perform optimally in differing noise environments. An exciting development is occurring outside the field of rehabilitation that will have a profound impact on the ability of the deaf to comprehend speech. Electronics companies are now beginning to market universal translation aids for travelers, where a phrase spoken in one language is captured, parsed, translated, and restated
(either spoken or displayed) in another language. The deaf would simply require that the visual display be in the language that they use for writing Deafness is often brought on (or occurs congenitally) by damage to the cochlea. The cochlea normally transduces variations in sound pressure intensity at a given frequency into patterns of neural discharge. TI neural code is then carried by the auditory (eighth cranial)nerve to the brainstem, where it is preprocessed and relayed to the auditory cortex for initial processing and on to the parietal and other cortical areas information extraction. Similar to the case for the visual system, the cochlea, auditory nerve, auditory cortex, and all relays in between maintain a topological map, this time based on tone frequency( tonotopic). If deafness is solely due to cochlear damage(as is often the case)and if the auditory nerve is still intact, a cochlear implant an often be substituted for the regular transducer array(the cochlea) while still sending the signal through he normal auditory channel ( to maintain modality specificity) At first glance, the design of a cochlear prosthesis to restore hearing appears daunting. The hearing range of a healthy young individual is 20 to 16,000 Hz. The transducing structure, the cochlea, has 3500 inner and 12,000 outer hair cells, each best activated by a specific frequency that causes a localized mechanical resonance in the basilar membrane of the cochlea. deflection of a hair cell causes the cell to fire an all-or pulsatile) neuronal discharge, whose rate of repetition depends to a first approximation on the amplitude of the stimulus. The outputs of these hair cells have an orderly convergence on the 30,000 to 40,000 fibers that make up the auditory portion of the eighth cranial nerve. These afferent fibers, in turn, go to brainstem neurons that process and relay the signals on to higher brain centers Klinke, 1983]. For many causes of deafness, the hair cells are destroyed, but the eighth nerve remains intact. Thus, if one could elicit activity in a specific output fiber by means other than the hair cell motion, perhaps some sense of hearing could be restored. The geometry of the cochlea helps in this regard as different portions of the nerve are closer to different parts of the cochlea Electrical stimulation is now used in the cochlear implant to bypass hair cell transduction mechanisms[ Loeb, 1985; Clark et al., 1990]. These sophisticated devices have required that complex signal processing, electronic, and packaging problems be solved One current cochlear implant has 22 stimulus sites along the scala tympani of the cochlea. Those sites provide excitation to the peripheral processes of the cells of the eighth cranial nerve, which are splayed out along the length of the scala. The electrode assembly itself has 22 ring electrodes spaced along its length and some additional guard rings between the active electrodes and the receiver to aid in securing the very flexible electrode assembly after it is snaked into the cochleas very small(a few millimeters)round window(a surgeon related to me that positioning the electrode was akin to pushing a piece of cooked spaghetti hrough a small hole at the end of a long tunnel). The electrode is attached to a receiver that is inlaid into a slot milled out of the temporal bone. The receiver contains circuitry that can select any electrode ring to be a source and any other electrode to be a sink for the stimulating current, and that can rapidly sequence between various pairs of electrodes. The receiver is powered and controlled by a radiofrequency link with an external transmitter, whose alignment is maintained by means of a permanent magnet imbedded in the receiver. igital signal processor stores information about a specific user and his or her optimal electrode locations for specific frequency bands. The object is to determine what pair of electrodes best produces the subjective perception of a certain pitch in the implanted individual himself or herself, and then to associate a particular lter with that pair via the controller. An enormous amount of compression occurs in taking the frequency range necessary for speech comprehension and reducing it to a few discrete channels. At present, the optimum ompression algorithm is unknown, and much fundamental research is being carried out in speech processing, compression, and recognition. But, what is amazing is that a number of totally deaf individuals can relearn to omprehend speech exceptionally well without speech-reading through the use of these implants. Other indi- viduals find that the implant aids in speech-reading. For some, only an awareness of environmental sounds is apparent; and for another group, the implant appears to have little effect. But, if you could(as I have been able to)finally converse in unaided speech with an individual who had been rendered totally blind and deaf by a traumatic brain injury, you would certainly begin to appreciate the power of rehabilitation engineering 119. 3 Engineering Concepts in Motor Rehabilitation Limitations in mobility can severely restrict the quality of life of an individual so affected. a wheelchair is a prime example of a prosthesis that can restore personal mobility to those who cannot walk. Given the proper c 2000 by CRC Press LLC
© 2000 by CRC Press LLC (either spoken or displayed) in another language. The deaf would simply require that the visual display be in the language that they use for writing. Deafness is often brought on (or occurs congenitally) by damage to the cochlea. The cochlea normally transduces variations in sound pressure intensity at a given frequency into patterns of neural discharge. This neural code is then carried by the auditory (eighth cranial) nerve to the brainstem, where it is preprocessed and relayed to the auditory cortex for initial processing and on to the parietal and other cortical areas for information extraction. Similar to the case for the visual system, the cochlea, auditory nerve, auditory cortex, and all relays in between maintain a topological map, this time based on tone frequency (tonotopic). If deafness is solely due to cochlear damage (as is often the case) and if the auditory nerve is still intact, a cochlear implant can often be substituted for the regular transducer array (the cochlea) while still sending the signal through the normal auditory channel (to maintain modality specificity). At first glance, the design of a cochlear prosthesis to restore hearing appears daunting. The hearing range of a healthy young individual is 20 to 16,000 Hz. The transducing structure, the cochlea, has 3500 inner and 12,000 outer hair cells, each best activated by a specific frequency that causes a localized mechanical resonance in the basilar membrane of the cochlea. Deflection of a hair cell causes the cell to fire an all-or-none (i.e., pulsatile) neuronal discharge, whose rate of repetition depends to a first approximation on the amplitude of the stimulus. The outputs of these hair cells have an orderly convergence on the 30,000 to 40,000 fibers that make up the auditory portion of the eighth cranial nerve. These afferent fibers, in turn, go to brainstem neurons that process and relay the signals on to higher brain centers [Klinke, 1983]. For many causes of deafness, the hair cells are destroyed, but the eighth nerve remains intact. Thus, if one could elicit activity in a specific output fiber by means other than the hair cell motion, perhaps some sense of hearing could be restored. The geometry of the cochlea helps in this regard as different portions of the nerve are closer to different parts of the cochlea. Electrical stimulation is now used in the cochlear implant to bypass hair cell transduction mechanisms [Loeb, 1985; Clark et al., 1990]. These sophisticated devices have required that complex signal processing, electronic, and packaging problems be solved. One current cochlear implant has 22 stimulus sites along the scala tympani of the cochlea. Those sites provide excitation to the peripheral processes of the cells of the eighth cranial nerve, which are splayed out along the length of the scala. The electrode assembly itself has 22 ring electrodes spaced along its length and some additional guard rings between the active electrodes and the receiver to aid in securing the very flexible electrode assembly after it is snaked into the cochlea’s very small (a few millimeters) round window (a surgeon related to me that positioning the electrode was akin to pushing a piece of cooked spaghetti through a small hole at the end of a long tunnel). The electrode is attached to a receiver that is inlaid into a slot milled out of the temporal bone. The receiver contains circuitry that can select any electrode ring to be a source and any other electrode to be a sink for the stimulating current, and that can rapidly sequence between various pairs of electrodes. The receiver is powered and controlled by a radiofrequency link with an external transmitter, whose alignment is maintained by means of a permanent magnet imbedded in the receiver. A digital signal processor stores information about a specific user and his or her optimal electrode locations for specific frequency bands. The object is to determine what pair of electrodes best produces the subjective perception of a certain pitch in the implanted individual himself or herself, and then to associate a particular filter with that pair via the controller. An enormous amount of compression occurs in taking the frequency range necessary for speech comprehension and reducing it to a few discrete channels. At present, the optimum compression algorithm is unknown, and much fundamental research is being carried out in speech processing, compression, and recognition. But, what is amazing is that a number of totally deaf individuals can relearn to comprehend speech exceptionally well without speech-reading through the use of these implants. Other individuals find that the implant aids in speech-reading. For some, only an awareness of environmental sounds is apparent; and for another group, the implant appears to have little effect. But, if you could (as I have been able to) finally converse in unaided speech with an individual who had been rendered totally blind and deaf by a traumatic brain injury, you would certainly begin to appreciate the power of rehabilitation engineering. 119.3 Engineering Concepts in Motor Rehabilitation Limitations in mobility can severely restrict the quality of life of an individual so affected. A wheelchair is a prime example of a prosthesis that can restore personal mobility to those who cannot walk. Given the proper