Disorders A - Z:   A    B   C    D    E    F    G    H    I    J    K    L    M    N    O    P    Q    R    S    T    U    V    W    X    Y    Z

Skip secondary menu

Neural Prosthesis Program Workshop


 

Table of Contents

Top

 

The Response of the CNS to Injury

Ohio State University, Phil Popovich
The goal of this study was to improve our understanding of the cellular response to injury in the spinal cord or brain. When a hemorrhage is introduced into the CNS there is resulting cell death and an inflammatory tissue response. The extent of this neuroinflammation determines the ability of the nervous system to regenerate. The degree of neuroinflammation is directly related to the extent of the vascular injury with injuries in the spinal cord resulting in greater damage than similar injuries in the brain. It was suggested that the level of vascular damage introduced by insertion of the microelectrodes would result in little if any neuroinflammation.

Top

 

Safe Stimulation of Neural Tissue

Huntington Med. Res. Inst., Doug McCreery, Al Lossinsky
Future neuroprosthetic devices propose the use of microelectrode arrays implanted in the CNS. This study addressed the effect of different stimulus parameters on inducing neural damage as well as stimulation-induced depression of neural excitability (SIDNE). The results suggest that stimulation with 6 nC/phase and 100% duty cycle does not generate neural damage, however increased charge injections (20-40 nC/phase) were damaging but the damage could be alleviated with decreased duty cycles. SIDNE became a factor when multiple contacts were used and interleaving the stimulation of the different contacts was not able to decrease this effect. It was suggested that blunt tip (6 mm tip radius of curvature) electrodes were good for current-density and microhemorrage considerations.

Top

 

Experiences with the Neurotrophic Electrode in Locked-in Humans

Neural Signals Inc., Phil Kennedy
Dr. Kennedy reported on the use of neurotrophic electrodes implanted in the motor cortex of a locked-in patient to control a cursor based on the recorded neural activity. Dissociation of facial muscle EMG and cortex controlled cursor movement was achieved with training. Phase of the neural signal was related to the direction of movement, whereas firing rate was not correlated with directionality. It was concluded that the success of the system in humans will depend on technological advances that make improvements to the signal to noise ratio. Rat and monkey histological studies indicate there is no neural damage resulting from insertion of the electrodes. Use of neurotrophic factors in the electrode is being explored to optimize the number of neurons residing inside the glass cone.

Top

 

A Tissue Engineering Approach To Understand Regeneration, Glial Scar and Interfaces

Case Western Reserve University, Ravi Bellamkonda
Recent work in neural regeneration has resulted in great advances toward bridging the gaps induced by injury and in understanding the mechanisms involved in neuronal growth both in the peripheral and central nervous system. However, several issues remain to be solved to promote neuronal growth into bridges and across glial scar tissue. Dr. Bellamkonda is studying the use of tissue engineering to solve these issues and to gain a better understanding of the mechanisms involved. He has created 3D gel matrices that incorporate haptotaxic and chemotaxic processes to promote neuronal cell growth and guided axonal extension. He has used layers of these gels to model the glial scar and explore methods for dissolving the glial scar and allow axonal growth.

Top

 

Brain Prostheses: Tissue Compatibility and Integration

Wadsworth Center, NY Dept. of Health, Bill Shain
The effect on tissue viability of single and multiple microelectrode devices inserted into the brain was evaluated using staining techniques to identify individual cells or cell products attached to the microelectrodes. The results indicate there are two distinct reactive responses: an early response characterized by inflammation and a prolonged response arising from cell interaction (astrocytes, microglia, endothelial cells) around the device. The size of the probe has an effect on the early response (diminished response with smaller probes), but the prolonged response is the same regardless of size. Understanding the mechanisms involved in generating these responses will help explain why some of these devices fail and should suggest methods for regulating the reactive response.

Top

 

Deep Brain Stimulation

University of Toronto, Andres Lozano
Deep brain stimulation (DBS) in structures of the thalamus is an effective treatment for a variety of movement disorders. However, little is know about the long-term effects of stimulation on the efficacy of the treatment as well as the extent of stimulation induced tissue damage. The results show little to no stimulation induced neural damage and for the majority of the patients, DBS remained an effective treatment over a 5 year period. However, some patients no longer needed the device after years of stimulation. Possible reasons for this phenomenon were a progressive thalamotomy from the stimulation, natural progression of the disease, or permanent physiological changes induced by the stimulation.

Top

 

Panel Discussion - What are the risks associated with CNS implantation and stimulation? What additional work needs to be done

Terry Hambrecht, Moderator
Neural damage associated with implanted electrodes can be attributed to the insertion of the electrode, the presence and micromovement of the electrode and stimulation through the electrode. The mechanisms involved in neural damage induced by electrical stimulation are not clear and they are probably related to either cell activation, passage of current through surrounding tissue or induced electric and magnetic fields or a combination of these events. Future work should examine new techniques to evaluate electrical vs. mechanical damage to better guide the development of microelectrodes for implantation in the CNS. Such research will probably lead to development of electrodes with materials that match the mechanical impedance of the brain and to treatments that can counteract the reactive response induced by the electrodes.

Top

 

Cortical Control of Neural Prostheses

Arizona State University, Andy Schwartz, Daryl Kipke
This group has been successful in obtaining stable single unit recording of the rat barrel cortex for up to four weeks in all the preparations and up to eight months in one preparation. Better results are expected with polyimide electrodes developed at ASU which are flexible and can be filled with neurotrophic factors to promote neuronal growth into the device. Consistent recordings were obtained from microwire electrode arrays implanted into the motor and pre-motor cortex of monkeys, allowing classification of task-related and directional cells. Monkeys were trained to control a virtual target based on the population vector obtained from the neuronal activity. Current work is exploring the feasibility of training to tune individual units for a particular direction and allow three-dimensional control using three perfectly tuned cells in orthogonal directions. Future work is aimed at controlling a robotic arm with the cortical signals.

Top

 

Cortical Control of Neural Prostheses

Brown University, John Donoghue, Kenji Nakata, Mijail Serruya
UTAH electrode arrays (100 channels, 74 active) were implanted into the arm area of the premotor cortex in monkeys. Neural activity recordings have been stable for over two years in one animal, with a 40% yield from the electrode arrays. Recordings made while the animal is performing two dimensional endpoint movements show that the neural units are directionally tuned. Offline reconstruction of the kinematics can be obtained using a linear filter that relates neural firing rate to endpoint position and velocity. The accuracy of the reconstruction depends on the length of the filter and the size of the neuronal ensemble. This approach will be used in online experiments to control the movement of a robotic arm.

Top

 

Direct Brain Interface Based on Event Detection in ECoG

University of Michigan, Simon Levine
The goal of this work is to create an interface with all types of assistive technology based on electrocorticogram (ECoG) recordings. Previous work has shown that ECoG recordings have better resolution than EEG, can be used in multi-channel control, have less artifact and can be correlated to events. Preliminary work with subjects who had subdural electrodes implanted in several different areas of the cortex has shown that high accuracy can be achieved even without training and that the ECoG best correlated to the motor task was recorded from the motor areas of the cortex. Future research will investigate imagined movements for control, correlation of brain activity with external measurements using fMRI and use of multiple channels to take advantage of the redundancy for accuracy improvement.

Top

 

Chronic Microstimulation of the Lumbosacral Spinal Cord

Huntington Med. Res. Inst., Doug McCreery, Al Lossinsky
Microelectrodes were implanted into the S2 region of the spinal cord of cats with the goal of activating the motoneurons of the urogenital system. Six working electrodes penetrated to the intermediate gray matter and two long stabilizer bars were used to keep the electrode array in place. High charge/phase resulted in stimulation induced neural damage, but this could be compensated for with low stimulation duty cycles. The use of microelectrode arrays is more difficult in the spinal cord than in the brain due to the increased movement of the electrodes resulting from body movement and tissue differences, as a result the stabilization bars are important. Spinal intact animals were used and these subjects showed no neurological deficits as a result of the presence of the implanted electrode array.

Top

 

Microstimulation of the Lumbosacral Spinal Cord - Mapping

University of Pittsburgh, Jim Roppolo
The goal of this study was to provide information on the location and distribution of specific neuronal populations within the spinal cord. Colon and anal sphincter pressures were measured as well as hind limb movements during microstimulation of the cat spinal cord. Stimulation in the intermediate gray area of S2 was able to generate colon pressures greater than sphincter pressures. Gait-like hind limb movements could be generated by stimulating in one spot in the lumbar region of the cord using sinusoidal variation of the stimulus amplitude. Flexion/extension as well as adduction/abduction of the limb were dependent on the electrode position and stimulus amplitude.

Top

 

Microstimulation of the Lumbosacral Spinal Cord - Mapping

Case Western Reserve University, Warren Grill, Michel Lemay, Ken Gustafson
This study focused on three areas important for the development of a neuroprosthesis based on microstimulation of spinal interneurons to generate distributed outputs. First, anatomical mapping was done to identify inhibitory interneurons of the urethral sphincter. Second, hindlimb motor responses were characterized by measuring hindlimb forces as a function of limb and electrode location in attempts to utilize the equilibrium point hypothesis. And third, computer modeling and fMRI techniques are being employed to develop methods to map the location of specific neuronal populations in the human spinal cord. It was suggested that one would need to activate four different interneurons controlling four different equilibrium points (on each side of the spinal cord) to generate robust controlled movement.

Top

 

Multichannel Stimulating and Intracortical Recording Electrodes

University of Michigan, Ken Wise
Thin film microprobes are commonly used for neuronal recording and stimulation. Unfortunately these probes suffer from a loss of signal over time. Attempts are being made to improve this limitation with different contact sizes, shapes and polymer coatings. "Puffer" probes are also being developed with flowmeters that allow for chemical delivery. However, the viscosity of the fluid may become an issue with the micropores that follow the shank of the probe, but they may be able to overcome problem with thermopneumatic pumps. One goal of the study is to develop fully functional active stimulating and recording arrays of up to 256 channels with incorporation of telemetry interfaces allowing control of the stimulation contact and stimulus waveform

Top

 

Insulating Biomaterials

InnerSea Technologies, Dave Edell
Most of the work in the study has been concentrated on protecting the CMOS integrated circuits by development of coatings and efficient cleaning of the substrates. Results suggest that substrates that are not clean lead to circuit failures. Fluorocarbon and silicon films are used for coatings. Coating of flat surfaces and wires is easier than coating of three-dimensional structures. Cleaning the substrates with CO2 has been found to be effective.

Top

 

Hermetic Packages and Feedthroughs.

The University of Michigan, Khalil Najafi
Any implantable neuroprosthetic device must have hermetic sealings that last for decades. The goal of this project was to develop hermetic sealing techniques specifically designed for small implantable devices. A glass-silicon package was used with a polysilicon binding layer. Testing with telemetry monitoring of the humidity inside the package suggests lifetimes of several decades, however there may be a problem with polysilicon dissolution at high temperatures. Therefore, new directions include using boron-doped coatings to protect the polysilicon from corrosion.

Top

 

Functional Neuromuscular Stimulation (FNS)

Case Western Reserve University, Pat Crago, Kevin Kilgore, Rich Lauer, Bob Kirsch
Biomechanical models of commonly performed tendon transfers in C5-C6 SCI individuals to restore voluntary or stimulated wrist extension have been developed to simulate surgical strategies that overcome current limitations in the functional outcome of the procedure. A study to evaluate the efficacy of command control algorithms for controlling the hand grasp system indicates that variable gain and rectification based algorithms are most effective. Cortical control (EEG) of the hand neuroprosthesis was evaluated in three subjects resulting in decreased performance in the GRT test as compared to other control methods, and no difference in ADL tasks. New research directions in FNS of the upper extremity research will lead to expansion of this technology to other SCI populations (C1-C4 and C7) as well as enhancement of the existing systems for the C5-C6 population.

Top

 

Neural Control Multichannel Implantable System

Case Western Reserve University, Hunter Peckham, Michael Keith
A second generation upper extremity neuroprosthesis is being developed that overcomes some of the limitations present in the current systems and includes improvements in areas identified by neuroprosthesis users. Such improvements include simplified donning of the controller, improved finger extension, overhead reach and bilateral function. Some of these issues have been addressed with the development of an implantable wrist angle transducer, implanting electrodes in the intrinsic muscles of the hand, implanting the triceps muscle and finally incorporating EMG of voluntary muscles as a control signal. These improvements should also allow extension of the hand neuroprosthesis to the C7 population. The challenge in this research will be to develop technology that is simple and thus less cumbersome for patients to use.

Top

 

Unassisted Standing by FNS

Case Western Reserve University, Ron Triolo
Currently, nine individuals with paraplegia have been implanted with the standing FNS system. The system allows subjects to reach while standing, make transfers and introduces some mobility. Large variability in functional outcome across subjects has been observed, attributed mostly to differences in subject size. Biomechanical modeling will be used in the future to address this issue and also to design the sensor-based control for the system and provide the subjects with the ability to control their posture. Previous work on adaptive recruitment mapping has used artificial neural networks to linearize the input/output properties of the system to allow better control. Trunk/spine stability and current muscle stimulation methods (reverse recruitment order) were identified as the main impediments to unassisted standing.

Top

 

Speech Processors for Auditory Prostheses

House Ear Institute, Bob Shannon
Auditory prostheses represent the most successful neuroprosthetic device to date, however the best patients still cannot perform as well as normal hearing individuals. Therefore, the goal of this study was to compare different speech processing techniques. The results suggest that there is no difference between the clarion and nucleus devices and that the optimal number of electrodes is eight using matched frequency input (instead of frequency compression, which is normally used). The study also found that one week of training, one hour/day, could significantly improve speech recognition.

Top

 

Speech Processors for Auditory Prostheses

Research Triangle Institute, Blake Wilson, Dewey Lawson
Possibilities for improvement in the functional outcome of auditory prosthesis include identification of the mechanisms underlying the variability in the outcome and closing of the gaps between the physiology of normal hearing and the function that implants can provide. The functional outcome of combined electric and acoustic stimulation (EAS) in a patient with low frequency residual hearing resulted in as good or better speech recognition compared to electrical or acoustic stimulation alone. Issues with tissue damage caused by the electrodes arise when these devices are implanted into individuals who have some residual hearing. More research is needed to evaluate the effect of implant depth on the extent of hair cell damage. There may be a trade off between lost hearing following deeper implants and improved long term hearing as the residual hearing diminishes with time.

Top

 

Neurophysiological Effects of Simulated Auditory Prosthesis Stimulation

University of Iowa, Paul Abbas, Charles Miller, Jay Rubinstein
This study is using both animal and computer models to better understand cochlear stimulation. In the animal studies single fiber refractory studies were conducted to address the input / output function of high frequency stimulation used in auditory prostheses. Measurements are also being made of the spatial extent of excitation using intraneural recordings, however this has proved difficult to interpret due to vague tonotopic arrangement in the auditory nerve. Computer simulations of the response of the auditory nerve to stimulation suggest that stochastic ion channel fluctuations can be important factors in matching the models to experimental data.

Top

 

Feasibility of a Central Nervous System Auditory Prosthesis

Huntington Med. Res. Inst. - House Ear Institute, Doug McCreery, Bob Shannon
In a subset of patients the auditory nerve is no longer a possible site of stimulation and in turn the goal of this project is to attempt to design an auditory prosthesis that stimulates the auditory nucleus. The proposed device uses both penetrating microelectrodes and electrodes placed on the surface of the auditory nucleus. They have decided to place the microelectrodes in the ventral nucleus based on innervations of the auditory nerve, however there are questions whether the total tonotopic map of the auditory nucleus can be accessed due to anatomical and surgical considerations.

Top

 

Effects of Remaining Hair Cells on Cochlear Implant Function

University of Iowa, Paul Abbas, Charles Miller, Jay Rubinstein
Similar to the other work presented by this group a combination of computer modeling and animal experimentation were used to explore the role of remaining hair cells on cochlear implant function. In the computer modeling section of the project they explored the role of hair cell synaptic activity in combination with stochastic ion channel representations and found that synaptic noise increased the dynamic range of the neuronal firing by desynchronization from the electrical sinusoidal input. The compound action potential of the auditory nerve was also recorded in normal and deafened cats. Lower thresholds were observed in the animals with remaining hair cells as well as improved temporal responses.

Top

 

Protective Effects of Patterned Electrical Stimulation

UCSF, Pat Leake
This study addressed the effects of electrical stimulation of the cochlea on neuronal survival in cats deafened at birth. The results show ganglion cell density decrease as a function of the duration of deafness. However, electrical stimulation caused in an increase in spiral ganglion cell survival compared to non-stimulated deafened animals, but the cell density was less than that found in non-deafened animals. The results also suggest that the longer the stimulation is used the better the cell survival, and that high frequency stimulation is beneficial.

Top

 

Protective Effects of Patterned Electrical Stimulation

University of Melbourne, Rob Shepherd
Previous research shows there is a direct relationship between spiral ganglion cell loss and duration of deafness. Experimental results indicate there is extensive cell loss within the first 12 months and only 2% cell survival after eight years. Deafness results in significant reduction of spontaneous cell activity, significant reduction in the size of the cochlear nucleus (50%), rapid atrophy of the cells during the first week, particularly in the ventral cochlear nucleus, as well as significant reduction of the synaptic density in the auditory midbrain. Function changes in cell activation following deafness includes little spontaneous activity, increased refractoriness, activation threshold and latency, and changes in the tonotopic organization to a more rudimentary representation. Future research will investigate application of electrical stimulation combined with neurotrophic factors to maintain cell density and functionality.

Top

 

Multichannel Transcutaneous Cortical Stimulation System

Illinois Institute of Technology, Phil Troyk
256-channel transcutaneous stimulation systems for the visual cortex with magnetic coupling for power have been designed and implemented. Each 256-channel module consists of four 64-channel submodules, each with its own power supply. The connector shell and inserter have been redesigned for practical wire attachment. Development of the 5 MHz transmitter technology has been completed and the transmitter is fully functional. The durability of the ceramic packaging has been tested in 90°C saline for over two years without failures. Future work will make these devices available for various applications.

Top

Last updated April 15, 2011