Development of Micro-channel Arrays

for Peripheral Nerve Recording

David J. Edell

, Ronald R. Riso

and Hugh Herr

Innersea Technology Inc., 1 DeAngelo Drive, Bedford, MA., U.S.A.

Biomechatronics Dept., Media Lab, Massachusetts Institute of Technology, 75 Amherst St., Cambridge, MA., U.S.A.

Keywords: Peripheral Nerve Interface, Micro-Channel Array, Prosthesis Control, Prosthesis Sensory.

Abstract: MicroTube Array (MTA) technology was developed to create an axon regeneration interface for exchanging

motor and sensory data with residual nerves. Future clinical application will include sensory-motor

transducers for individuals with limb amputation. In this pilot study, a small matrix (9) of MTAs 1, 3 and

5mm long with either 99um, 200um or 287um diameter MicroTubes (MTs) filling nerve cuffs of 3mm in

diameter were implanted in tibial nerve of NZW rabbits and histologically evaluated after more than 6

months post-op. Full regeneration was observed in all 3 lengths for 287um MTAs, and for all three

diameters of MTs with 1mm length. The remaining implants were mechanically dislodged during the

healing phase. A second implant set was designed to include 12 platinum-iridium wire electrodes direct

wired to a percutaneous connector. Successful recordings of useful amplitudes were observed during reflex

righting behaviour for over 2 years before the anticipated wire breakage ended the experiments.

1 INTRODUCTION

Recent advances in the design of upper and lower

extremity prosthetic limbs has underscored the need

to provide more effective means for amputees to

control these prostheses as well as to provide

sensation from these devices. It is widely

appreciated that these goals could best be met by

establishing a permanent electrical interface with the

trunk nerves in an amputee’s residual limb (Mannard

et al. 1974; Edell 1986; Riso 1999), however the

development of the technology to achieve this has

been elusive. Among the approaches that have been

advocated and continue to be researched are: diverse

designs of nerve cuffs (Loeb and Peck 1996; Naples

et al. 1988; Walter et al. 1997; Grill and Mortimer

1998; Hoffer and Kallesoe 2000; Schuettler and

Stieglitz 2000), intrafasicularly placed fine wire

(Lefurge et al. 1991) and conductive polymer

filaments (McNaughton and Horch 1996; Lawrence

et al. 2002; Lawrence et al. 2004; Boretius et al.

2010), transverse penetrating arrays of

microfabricated needle electrodes [e.g. Utah “Slant

Array” (Clark et al. 2011; Wark et al. 2013)] and

“sieve” (Kovacs et al. 1992; Bradly et al. 1997;

Wallman et al. 2001; Lago et al. 2005) styled

devices. While progress continues to be made (see

for example Tan et al. 2014), none of these device

designs has been shown to completely meet the

needs of the prosthesis application in terms of

numbers of independent recording channels for

obtaining motor commands, provision for activating

discrete sensory afferents for feedback of tactile and

proprioceptive events, or device longevity (although

cuff designs have been successfully deployed for

other neuroprosthesis applications such as bowel and

bladder control (Creasey et al. 2001) or FES based

standing and walking for paraplegia (Schiefer et al.

2013).

The most useful device for an individual with

amputation would provide sufficient information

exchange between the nerve and prosthesis to enable

return of complete and natural sensorimotor

function. One approach is to embed the distal end of

the residual nerve within a nerve cuff that contains

properly designed and constructed MicroTube

Arrays (MTAs). MTAs are small diameter Micro

Tubes (MTs) that each contain neural recording and

activation functionality for use in motor control and

sensation. Under appropriate conditions, axons in all

severed nerves will regenerate into and through

small openings in MT devices.

In 1974, it was known that amphibian peripheral

J. Edell D., R. Riso R. and Herr H..

Development of Micro-channel Arrays for Peripheral Nerve Recording.

DOI: 10.5220/0005091500050012

In Proceedings of the 2nd International Congress on Neurotechnology, Electronics and Informatics (NEUROTECHNIX-2014), pages 5-12

ISBN: 978-989-758-056-7

 2014 SCITEPRESS (Science and Technology Publications, Lda.)

nerves would regenerate through small holes in

implant structures from the work of Mannard and

Stein (1974; Stein et al. 1975), but that approach was

not successful when applied to mammals. In 1980

Edell showed that with biocompatible materials and

design, the regeneration approach would work in

rabbit peripheral nerves (Edell 1980; Edell et al.

1982). However, the selectivity of the recordings

was limited by the open environment about the

electrodes. In 1977, Loeb et al. (1977) published a

method for isolating small groups of axons and

improving signals in peripheral nerves by having

them regenerate through MicroTubes, but was

unsuccessful in making it work. The theory was

sound - small diameter tubes increase signal

amplitudes by increasing the effective extracellular

resistance (Fitzgerald et al. 2008). However, the

choice of dimensions (~10x15um) for the tubes was

perhaps smaller than needed for the required support

cells, collagen and capillaries, and the materials’

surfaces may have triggered tissue micro-

incompatibility.

This paper describes previously unpublished

studies (DARPA project, 2006) demonstrating that

MTAs can be made to work in rabbit peripheral

nerves. Results are reported concerning an initial

study that surveyed the effects of MT diameter and

MT length on the ability of nerve fibers to grow into

such devices and provide stabile recordable, multi-

channel neural activity for use in prosthesis control.

Robust regeneration into 100um diameter MTs was

documented, the smallest diameter tested.

Histological evidence suggests that MTs as small as

25um may be possible if properly designed. Within

each MT there must be room for collagen, Schwann

cells, and capillaries for mechanical, axonal, and

metabolic support. These early findings have been

corroborated in recent studies in other laboratories

where successful neural regeneration has been

reported (in rodents) into devices having channel

sizes with cross sections of 100um x100um (Lacour

et al. 2009) or as small as 70um x 20um rectangles

(Stoyanov et al. 2013) and this results in the

subdivision of a composite nerve into mini-fascicles

which may enhance the separability of targeted

nerve fibers by functional type.

2 METHODS

2.1 MicroTube Array Device

Fabrication

MTAs were constructed within a Class 100

cleanroom from micro-polyimide tubing (Micro-

Lumen) aggregated together within 2.5mm ID

polyimide tubing. The interior diameters of the MTs

were either 287, 203 or 99um (Fig 1). The bundled

tubing was sliced transversely to produce micro-

channel arrays having defined lengths of 1, 2 and

3mm.

Figure 1: Top – photograph showing the microchannel

arrays that were constructed using three different lengths

of MTs. MTAs can be seen as the darken areas within the

nerve cuffs); Lower - Photographs showing the cross

sections of arrays that were comprized of the largest

diameter tubes (287um dia.) and the smallest diameter

tubes (99um dia.) used in this study. These arrays

contained approximately 40 and 400 MTs, respectively.

2.2 Incorporation of Recording

Electrodes into MicroTube Arrays

Figure 2: Magnified view of 3mm long, 203um diameter

MicroTubes (brown) with 50um diameter insulated

iridium microelectrode shafts placed halfway into 12 of

the MicroTubes. The portal visible at the center of the

picture is the exit port required for axon growth. Reference

electrode coils and nerve cuffs were added to both the

entry and exit chambers after this photo was obtained).

NEUROTECHNIX2014-InternationalCongressonNeurotechnology,ElectronicsandInformatics

Fig. 2 shows a photograph of an assembled MTA

nerve interface device in which 12 of the MTs are

instrumented for neural recording and/or

stimulation.

2.3 Experimental Design for in-Vivo

Studies

The first part of this pilot study was aimed at

determining the extent to which channel size (i.e.

cross sectional area) would influence the ability of

nerve fibers from a newly transected nerve to enter

and grow through the channels. A second objective

was to determine any effects of the channel length

on neural ingrowth. To address these issues, nine

New Zealand White rabbits (2.4-3kg) were

implanted with a nerve interface device from a

matrix of 9 devices representing combinations of the

three MT sizes (287, 203 and 99um dia.) and three

channel lengths (1, 3 and 5mm).

As this study was intended only as an initial

survey, only one animal could be assigned to each of

the nine permutations of channel size and channel

length. All of the animal studies were reviewed and

approved by the MIT- IACUC. With the animal

fully anesthetized and using sterile technique, the

sciatic nerve on one leg was exposed just proximal

to the knee, and the tibial nerve component was

isolated and transected taking care not to injure the

peroneal and sural nerve components. The proximal

end of the isolated tibial nerve was inserted into the

MTA device and secured using microsutures applied

through the epineurium. The distal segment of the

transected nerve was then inserted into the distal

opening of the nerve interface.

While this animal model is not a full amputation

model, it was selected because it allows for

monitoring of the progress of the nerve regeneration.

The experimenter can readily observe when the

regenerating nerve has traversed the nerve interface

device and reconnected to the distal nerve target

tissues. Thus, any return of ankle extensor function,

either volitional or reflexive, denotes successful

motor nerve regeneration through the device.

Sensory nerve regeneration can be accessed by

applying toe pinch and noting a return of flexor

withdrawal of the limb.

Activity can be conveniently elicited by placing

the animal prone on a table and gently rolling the

head and shoulders from side to side. Leg extension

will ensue as the animal acts to maintain balance. A

second technique that was used to evoke activity in

the ankle extensor muscle (and thus tibial nerve

activation) involves the “drop reflex” whereby the

animal is held in the air and then lowered towards

the table so that the foot extends in anticipation of

falling.

3 RESULTS

3.1 Assessment of Nerve Regeneration

After the tibial nerve regeneration was relatively

stable as judged by the return of voluntary and reflex

control of ankle extension function, each animal was

euthanized and the devices retrieved for histological

examination. Figure 3 shows the results seen with

one of the implants (99um channels and 1mm

length). The tissue was fixed in formalin and then

the attached nerve was pulled out of the interface

device, embedded in paraffin and stained. Not all of

the fine micro-fascicles would slide out of the

MicroTubes. However, for the most part, the micro-

channel array pattern is evident.

Figure 3: Left – Photograph of explanted 1mm long and

99um diameter MicroTube array assembly after

regenerated peripheral nerve was removed. Green circle

depicts the region from which the nerve mini-fascicles

shown at the right were withdrawn. Right – Micrograph

depicting the core of neural tissue that had grown through

the micro-channel nerve interface.

Among the 9 implanted animals there were three

mechanical failures most likely caused by fixation of

the implants by connective tissue (which occurs

rapidly) in a location that was not ideal

mechanically. Straightening of the leg, perhaps

when ear scratching, or during cage changes during

the early healing phases could have generated suffic-

ient stress on the interface to result in the nerve

pulling out. In one instance (5mm long, 203um

diameter array), partial regeneration occurred

through a collapsed cuff where the distal nerve had

been pulled out but the sutures held. The sutures

collapsed the cuff so only a small opening was

available for nerve regeneration to traverse on the

distal side. There was good histology obtained from

DevelopmentofMicro-channelArraysforPeripheralNerveRecording

the small subset of MTs that could support

regeneration in the limited space.

Most of the MTs were filled with myelinated and

(probably) non-myelinated axons, collagen and

fibroblasts, and most importantly, capillaries (see

summary in Table 1).

Table 1: Matrix of MT diameters and lengths studied.

Length\Dia 287 µm 203 µm 99 µm

1 mm Regeneration Regeneration Regeneration

3 mm Regeneration Pulled Out? Pulled Out?

5 mm Regeneration Partial* Pulled Out?

3.2 Speculation regarding Minimal

Micro-Channel Sizes

Figure 4: Upper – Theoretical graph showing the effect of

micro-channel diameter for a 3mm long MT vs. the

expected signal amplitude recordable from a single node

of Ranvier positioned at the center of a MT with either a

larger or smaller diameter. Lower – micrograph showing

regenerated neural tissue that was removed from one MT

of an explanted 3mm long and 99um dia. MTA assembly.

Capillaries can be seen in all “sampled” regions delineated

by the 33um dia. superimposed circles.

Figure 4 (upper) shows a graph of the theoretical

signal amplitude that could be recorded from one

node of Ranvier that is centered in 3mm long MTs

of different diameters. The micrograph (Fig. 4

lower) showing the neural tissue that regenerated

into a 99um dia. tube, was overlayed with small

(yellow) circles having 33um dia., and it can be seen

that there is a sufficient density of axons, capillaries

and support cells within each drawn circle to suggest

that regeneration would be successful for MT

diameters of this small size (33um). Notably, if this

regeneration were successful, then the use of small

MTs on the order of the 33um dia. could be expected

to afford axon signals in the range of 8mV as

indicated by the green arrow in the figure.

3.3 Recordings of Neural Activity

Based on results from the regeneration studies with

the 203um diameter x 3mm long MTA, and the fact

that the effects of the presence of 50um wires within

the tubes was unknown, we decided to use these

intermediate sized tubes as the basis for developing

the neural interface array that incorporated integral

electrodes.

Three devices were constructed and implanted.

Each of the implants was successful. The first two

animals showed signs of regaining muscle function,

and yielded good recordings after 6 weeks. The

third animal showed promising recordings after 4

weeks, but shortly thereafter the percutaneous

connector that was used to access the electrodes

failed. Useful recordings were able to be obtained

throughout the 7 months post implantation that the

animals were studied. Examples of recorded neural

activity from the first animal are shown in Figs. 5, 6

and 7.

Figure 5: Example waveforms from rabbit R06015 approx.

3 months post-implantation. While immature, axons were

beginning to exhibit robust responses to the righting reflex

(~6-13s. middle time section of graph) and steady

resistance to manipulation (~13.5s-end). (Max ampl. ~

180uV).

NEUROTECHNIX2014-InternationalCongressonNeurotechnology,ElectronicsandInformatics

For this animal, six of the eleven recording channels

yielded high quality signals (Fig. 5) while four

channels produced lower amplitude signals and one

channel had a failed lead at the time of implantation.

The capacity for selective channel recording was

clearly evident as there are frequent occurrences of

activity present on one channel that does not occur

on neighboring channels. In addition to

demonstrating freedom from channel crosstalk, this

shows that the recorded activity originates with

axons contained within each specific tube and is not

due to contaminating artifacts such as EMG from

muscle tissue that’s surrounds the implant (since

such activity would be expected to contaminate all

of the channels simultaneously if not excluded by

the differential recording instrumentation). There is

also a considerable amount of nerve activity that

occurs approximately at the same time among

different channels (Fig. 5). This is due to recruitment

of motor units or afferent feedback that is synergistic

since the neural waveforms are not coincident as

would be expected if they were from the same

neurons.

The independence of the recording channels in

this same animal is more readily apparent in Fig. 6

which shows two different epochs of nerve activity

(left and right columns) that were recorded

simultaneously using electrodes #3 and #8 (compare

upper vs. lower panels).

Figure 6: Example of recorded neural activity from animal

R06029 electrodes 3 and 8. LEFT - simultaneous

recordings showing distinct differences in activity levels

and type. RIGHT - recordings from another episode of

activity again showing little correlation other than similar

timing of general heightened activity. Data acquired 6

weeks post-implant.

3.4 Qualities of the Recorded Axons

from Regenerated, Maturing Axons

The recorded nerve activity showed a wide range of

action potential amplitudes as well as spike

durations. It isn’t known at this point why some

channels yielded higher amplitude recordings than

others. We can offer several speculations to explain

these effects: 1) Possibly the amplitude disparity is

due to the nerve fibers not being fully regenerated in

one channel verses another, as newly regenerated

fibers are known to have smaller diameters, and

thinner myelination than more mature fibers and

thus have lower amplitude action potentials. 2) With

regard to the range of action potential durations -

The design of the microtube constricts the small

axonal currents so that large signal amplitudes are

recorded from non-myelinated fibers as well as

myelinated fibers though the time course of the

waveforms are markedly different. It is worth

noting that the ability to record from non-myelinated

fibers may be desirable for neurophysiology studies

as small caliber autonomic fibers are normally

difficult to record. The small diameters, thin myelin,

and closely spaced nodes of the recently regenerated

axons are associated with slow nerve conduction

velocities and the action potentials from small fibers

(and particularly from unmyelinated immature fibers

or autonomic nerve fibers) tend to be of long

duration. 3) Some of the largest recorded activity

may represent compound action potentials formed

from the superposition of quasi-simultaneous nerve

discharges within the tube structures. 4) There is a

wide distribution of fiber diameters in any peripheral

nerve so that even a normal nerve has a wide

distribution of amplitudes and action potential

duration due to the fiber diameter distributions.

3.5

Utility of Combined Activity from

Different Channels

Figure 7: Composite signal from combining channels

1,2,3,9,10,11 yields a more predicable response than

single channels due to the effective increase in motor unit

pool being recorded. Note pulsed responses due to

rhythmic righting reflex and relatively steady amplitude

due to continuous force on leg, followed by relatively

vigorous response at end. Amplitude bar: lower trace

approximately 85uV; upper trace was a smoothed/rectified

image of lower trace where the amplitude bar is about

8.5uV.

There have been several instances where

recordings of ‘whole nerve’ have been useful for

neuro-prosthetic devices (Haugland and Sinkjaer

1999; Sinkjaer et al. 1999; Riso and Slot 1996).

Using the microtube array it is possible to select

various combinations of channels to create the

DevelopmentofMicro-channelArraysforPeripheralNerveRecording

highest quality composite signals. Fig. 7 shows an

example of combining channels 1,2,3,9,10, and 11

from one experiment to achieve such a composite

signal to

better represent the intended contraction of a

particular tibial n. innervated muscle or group of

muscles than each waveform alone.

An estimate of the composite signal intensity

(solid line) is shown above the raw neural

waveforms. With smaller tubes, more ‘muscle

specific’ reconstruction of the efferent activity could

be achieved. In addition, tubes that contained

afferent nerve fibers could be designated for

stimulation for cutaneous or proprioceptive

sensation. Since axons tend to aggregate by function

in peripheral nerves, this approach should be robust.

4 DISCUSSION AND

CONCLUSIONS

4.1 Choice of Animal Model

Rabbits have long been accepted as adequate models

for peripheral nerve repair work in humans. Rats

and lower animals, particularly amphibians,

regenerate more robustly than rabbits and higher

animals, and thus, the results obtained with those

preparations are not as readily extrapolated to the

human condition. Because the studies presented

herein were performed using an animal model in

which the nerve interface served as a bridge between

the transected ends of the tibial nerve, for

experimental expediency, it might be argued that the

successful regeneration of the nerve into the device

may depend strongly on the ability of the proximal

nerve to reconnect to the distal portion of the

transected nerve and thus may not be applicable for

interfacing to the truncated nerves in the amputee’s

residual limb.

At issue here is not whether the transected nerve

will regenerate into a MTA. So long as the distal end

of the MTA remains sufficiently open, the nerve

fibers will exit the device and form into a

disorganize mass as a neuroma. It is generally best,

however, to avoid the occurrence of a neuroma

because neuroma tissue frequently is overly

sensitive to mechanical stimulation. This can result

in painful sensations unless specific tactics are used

to shield the neuroma from such stimulation.

One solution to prevent the formation of a

neuroma in the absence of the distal nerve segment

for attachment, is to provide small pieces of target

tissue (e.g. muscle or skin) at the distal exit of the

nerve interface. This target tissue allows the

regenerated nerve fibers to terminate onto

appropriate end organs and has a stabilizing effect

on the nerve. A possible contra-indication with

applying this strategy, however, is that the receptors

present in the target tissue may generate activity in

the connected nerve that could conflict with the

intended nerve recording or stimulation paradigm.

Further research in this area is warranted.

4.2 Importance of Channel Size

Smaller (more narrow) channels favor improved

separability of axons so that efferent fibers to

muscles that normally would control different

functions don’t co-mingle within the same channel.

Additionally, segregation of the motor units

associated with any given muscle into smaller

groups allows for finer resolution for the derived

control signal. Where desired, perhaps to achieve a

more robust control signal, the activity of those

channels could always be recombined. Furthermore

as previously stated, the recorded signal amplitude is

larger for smaller diameter channels.

With regard to electrical stimulation to provide

sensory feedback, it would be best to be able to

selectively activate afferent fibers that shared a

common modality and region of referred sensation.

While there is some evidence from

microneurography studies in man (Stoyanov et al.

2013; Ochoa and Torebjörk 1983), that suggests

grouping of sensory axons in peripheral nerves by

modality, and end receptor location, presently too

little is known about the extent of such grouping as

they course, for example, from the brachial plexus,

down the arm, and to the wrist and fingers to be able

to suggest optimal channel sizes.

Ultimately, selection of an “optimal” channel

size might depend on the intended application for

the nerve interface and the particular morphology of

the targeted nerve. The best strategy currently

would seem to be to develop a modular system that

would be scalable and adaptable to accommodate

different size nerves and different locations within

the amputee’s residual limb. Large trunk nerves are

intrinsically divided into small fascicles by

perineurium, and within the small fascicles, axons

aggregate mostly by function. As the axons near

their targets, they further segregate into individual

fascicles before exiting the nerve trunk. With a

scalable system, the level of amputation would not

present significant difficulty as it is simple to deploy

individual interface devices to each of the divided

NEUROTECHNIX2014-InternationalCongressonNeurotechnology,ElectronicsandInformatics

nerve fascicles. Most fascicles are 0.5-2mm

diameter. Some fascicles are larger, but can easily

be divided as needed to accommodate a standard

MTA interface sub-unit of perhaps 2.5mm diameter.

An additional benefit of employing the micro-

channel approach to nerve interfacing is that the

same micro-channel could be used for both motor

and sensory functions as desired. This feature may

be particularly important for groups of axons

destined for a muscle, as motor nerves carry a great

deal of sensory information about the muscle target.

Once the physiological function of each micro-

fascicle is determined, information can be combined

as appropriate.

4.3 Design Considerations for MTA

Devices

It should be noted that in spite of its use in this pilot

work, polyimide is NOT a suitable long term

implant material – it undergoes slow hydrolytic

degradation and will over time fail as an insulator.

Further, the wall thickness is far too great for use in

a good clinical device. The open area occluded by

wall thickness can completely prevent regeneration.

While in these simple single fascicle experiments the

open area of the devices can be kept to at least the

open area of the fascicle, in a practical device for

human shoulder, arm or leg amputation applications

might require many of these devices and result in

unnecessary bulk. A better solution would be to use

silicon dioxide or titanium dioxide MTs fabricated

on low-power integrated circuit substrates. Such

substrates would assist to mechanically stabilize the

nerve interface, and the integrated electronics could

be insulated reliably (Edell 2004) to survive

implantation for decades with minimal bulk.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the

assistance of the Huntington Medical Research

Laboratory for providing the Platinum-Iridium

microwire electrodes that were used in developing

the recording aspect of the nerve interface design.

This research was supported by: DARPA grant

DSO N66001-05-C-8030, “Peripheral Nerve

Interface Technology for Bidirectional Neural

Communication” H. Herr-PI, with sub-contract to

InnerSea Tech. Inc.

REFERENCES

Boretius T, Badia J, Pascual-Font A, Schuettler M,

Navarro X, Yoshida K and Stieglitz T 2010, ‘A

transverse intrafascicular multichannel electrode

(TIME) to interface with the peripheral nerve’,

Biosens. Bioelectron., vol. 26, pp. 62–69.

Bradley RM, Cao X, Akin T and Najafi K 1997, ’Long

term chronic recordings from peripheral sensory fibers

using a sieve electrode array’, J. Neurosci. Methods,

1997, vol. 73, no. 2, pp. 177-186.

Clark GA, Ledbetter NM, Warren DJ and Harrison RR

2011, ‘Recording sensory and motor information from

peripheral nerves with Utah Slanted Electrode Arrays’,

in Proceedings of the 33rd Annual International

Conference of the IEEE Engineering in Medicine and

Biology Society, Aug 30-Sep 3, Boston pp. 4641-4644.

Creasey GH, Grill W, Korsten M, Sang HU, Betz R.

Anderson R and Walter J 2001, ‘An implantable

neuroprosthesis for restoring bladder and bowel

control to patients with spinal cord injuries: A

multicenter trial,’ Arch. of Physical Med. and Rehab.

vol. 82, no. 11, pp. 1512–1519.

Edell DJ 2004 ‘Insulating Biomaterials’, Chapter 3.2, in

Neuro-prosthetics Theory and Practice, eds KW

Horch and GS Dhillon, World Scientific Publishing

Co., pp. 517-579.

Edell DJ 1980, Dissertation: ‘Development of a Chronic

Neuroelectric Interface’, University of California.

Edell DJ, Churchill JN and Gourley IM 1982, ‘Biocom-

patibility of a silicon based peripheral nerve

electrode”, Biomat. Med. Dev. Art. Org., vol. 10, pp.

103-122.

Edell DJ 1986 ‘A peripheral neural information transducer

for amputees: long- term multichannel recordings from

rabbit peripheral nerves’, IEEE Trans. Biomed. Eng.,

vol. 33, pp. 203-214.

Fitzgerald JJ, Lacour SP, McMahon SB and Fawcett JW

2008 ’Microchannels as axonal amplifiers’, IEEE

Transactions on Biomedical Engineering , vol. 55, no.

3, pp. 1136-1146.

Grill WM and Mortimer JT 1998,’Stability of the input

output properties of chronically implanted multiple

contact nerve cuff stimulating electrodes’, vol. 6, no.4,

pp. 364-373.

Haugland M and Sinkjær T, ‘Interfacing the body's own

sensing receptors into neural prosthesis devices,’

Technology and Health Care, vol. 7, no. 6, pp.393-

399, 1999.

Hoffer JA and Kallesoe K 2000, ‘How to use nerve cuffs

to stimulate, record or modulate neural activity’.

Chapter 5, In: Neural Prostheses for Restoration of

Sensory and Motor Function, K.A. Moxon and J.K.

Chapin Eds., CRC Press, pp. 139-175.

Kovacs GTA, Storment CW and Rosen JM 1992,

‘Regeneration microelectrode array for peripheral

nerve recording and stimulation’, IEEE Trans.

Biomed. Eng., vol. 39, pp. 893-902.

Lacour SP, Fitzgerald JJ, Lago N, Tarte E, McMahon S

and Fawcett J 2009, ‘Long Micro-Channel Electrode

DevelopmentofMicro-channelArraysforPeripheralNerveRecording

Arrays: A Novel Type of Regenerative Peripheral

Nerve Interface’, IEEE Trans. on Neural Systems and

Rehab. Eng., vol. 17, no.5, pp. 464-460.

Lago N, Ceballos D, Rodríguez FJ, Stieglitz T, and

Navarro X 2005, ‘Long term assessment of axonal

regeneration through polyimide regenerative

electrodes to interface the peripheral nerve’,

Biomaterials, 2005, vol. 26, no.14, pp. 2021-2031.

Lawrence SM, Larsen JO, Horch KW, Riso R, and

Sinkjear T, 2002, ‘Long-Term biocompatibility of

implanted polymer-based intra-fascicular electrodes,’

J. Biomed. Mat. Res., 2002, vol. 63, pp. 501-506.

Lawrence SM, Dhillon GS, Jensen W, Yoshida K and

Horch KW 2004, ‘Acute peripheral nerve recording

characteristics of polymer-based longitudinal

intrafascicular electrodes’, IEEE Trans. Neural. Syst.

Rehabil. Eng., vol. 12, no. 3, pp. 345-348.

Lefurge T, Goodall E, Horch K, Stensaas L and

Schoenberg A 1991, ‘Chronically implanted

intrafascicular recording electrodes’, Annals of

Biomedical Engineering, vol. 19, no. 2, pp. 197-207.

Loeb GE, Marks WB and Beatty PG 1977, ‘Analysis and

microelectronic design of tubular electrode arrays

intended for chronic, multiple single-unit recording

from captured nerve fibres’, Med. Biol. Eng. Comput.,

1977, vol. 15, pp. 195-201.

Loeb GE and Peck RA 1996, ‘Cuff electrodes for chronic

stimulation and recording of peripheral nerve activity’,

J Neurosci Methods, vol. 64, no. 1, pp. 95-103.

Mannard A, Stein RB and Charles D 1974, ‘Regeneration

electrode units: implants for recording from single

peripheral nerve fibers in freely moving animals’,

Science, vol. 183, no. 124, pp. 547-549.

McNaughton TG and Horch KW 1996 ‘Metallized

polymer fibers as leadwires and intrafascicular

microelectrodes’, J. Neurosci. Methods, vol. 70, no.1,

pp. 103-110.

Milkman, R 1998, 'The new American workplace:high

road or low road?' in Workplaces of the future, eds P

Thompson and C Warhurst, Macmillan Press, London,

pp. 22-34.

Naples GG, Mortimer JT, Scheiner A, and Sweeney JD

1988, ‘A spiral nerve cuff electrode for peripheral

nerve stimulation’, IEEE Trans. Biomed. Eng., vol. 35,

pp. 905-916.

Ochoa J and Torebjörk E 1983, ‘Sensations evoked by

intraneural micro-stimulation of single mechano-

receptor units innervating the human hand’, J. Physiol.

vol. 342, pp. 633-654.

Ochoa J and Torebjörk E 1989, ‘Sensations evoked by

intraneural micro-stimulation of C nociceptor fibres in

human skin nerves,’ J Physiol. vol. 415, pp. 583-599.

Riso RR and Slot PJ 1996, ‘Characterization of the ENG

activity from a digital nerve for feedback control in

grasp neuroprostheses’, in Neuroprosthetics: from

Basic Research to Clinical Applications. eds A

Pedotti, M Ferrarin, J Quintern, R Riener, Springer,

Berlin, pp. 345-357.

Riso RR, 1999, ‘Strategies for providing upper extremity

amputees with tactile and hand position feedback-

moving closer to the bionic arm’, Technology in

Health Care, vol. 7, no. 6, pp. 401-409.

Schiefer MA, Freeberg M, Pinault GJ, Anderson J, Hoyen

H, Tyler DJ and Triolo RJ 2013,‘Selective activation

of the human tibial and common peroneal nerves with

a flat interface nerve electrode’, J Neural Eng., vol.

10, no.5, 056006. PubMed PMID: 23918148; PubMed

Central PMCID: PMC3809099.

Schuettler M and Stieglitz T 2000, ‘18polar hybrid cuff

electrodes for stimulation of peripheral nerves’, in:

Proceedings of the 5th Annual Conference of the

International Functional Electrical Stimulation

Society, Alborg, Denmark, pp. 265-268.

Sinkjær T, Haugland M, Struijk J, and Riso RR 1999,

‘Long-term cuff electrode recordings from peripheral

nerves in animals and humans’, in Modern Techniques

in Neuroscience Research. eds U Windhorst and H

Johansson, Springer, Berlin, pp. 787-802.

Stein RB, Charles D, Davis L, Jhamandas J, Mannard A,

and Nichols TR 1975 ‘Principles underlying new

methods for chronic neural recording.’ Can J Neurol

Sci, vol. 2, pp. 235-244.

Stoyanov II, van Wezel RJ and Rutten WL 2013, ‘In vivo

testing of a 3D bifurcating microchannel scaffold

inducing separation of regenerating axon bundles in

peripheral nerves’, J Neural Eng., vol. 10, no. 6,

066018 (13pp)

Tan D, Schiefer M, Keith MW, Anderson R and Tyler DJ

2013, ‘Stability and selectivity of a chronic, multi-

contact cuff electrode for sensory stimulation in a

human amputee’, in Proceedings of the 6

International Conference on Neural Engineering

(NER IEEE/EMBS), Nov 6-8, San Diego, CA, pp. 859-

862.

Wallman L, Zhang Y, Laurell T and Danielsen N 2001,

‘The geometric design of micromachined silicon sieve

electrodes influences functional nerve regeneration’,

Biomaterials, 2001. vol. 22, no. 10, pp. 1187-1193.

Walter JS, Griffith P, Sweeney J, Scarpine V, Bidnar M,

McLane J and Robinson C 1997, ‘Multielectrode

nerve cuff stimulation of the median nerve produces

selective movements in a raccoon animal model’, J

Spinal Cord Med., vol. 20, no. 2, pp.233-243.

Wark HA, Sharma R, Mathews KS, Fernandez E, Yoo J,

Christensen B, Tresco P, Rieth L, Solzbacher F,

Normann RA and Tathireddy P, 2013, ‘A new high-

density (25electrodes/mm²) penetrating microelectrode

array for recording and stimulating sub-millimeter

neuroanatomical structures’, J Neural Eng.,vol. 10, no.

4, pp.

NEUROTECHNIX2014-InternationalCongressonNeurotechnology,ElectronicsandInformatics