This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
1
The fabrication, implantation and stability of
intraspinal microwire arrays in the spinal cord
of cat and rat
Jeremy A. Bamford, R. Marc Lebel, Kian Parseyan and Vivian K. Mushahwar, Member, IEEE
Abstract—Intraspinal microstimulation (ISMS) is currently
under investigation for its ability to restore function following
spinal cord injury and aid in addressing basic investigations of the
spinal cord in feline and murine (rat) models. In this report we
describe the procedures for fabricating and implanting intraspinal
microwires, with special emphasis on the rat model. Furthermore,
we report our results on targeting success and long-term stability
and functionality of the implants. Early targeting with implants
fabricated based on general ‘average’ dimensions of the spinal
cord was approximately 50% successful in reaching the proper
targets within the ventral grey matter in cats. Improvements in
insertion tech- nique and the use of multiple contact electrodes
have raised the targeting success to 100%. Furthermore, the
manufacturing of ISMS arrays has been improved by the use of
magnetic resonance imaging to create subject-specific implants for
cats and track the location of the arrays post-implant. In the rat,
our procedures have produced desirable targeting of all recovered
microwires. We speculate this is due to the different targeting
parameters and the shorter depth of insertion in the rat spinal
cord. Although there is a heightened mechanical mismatch
between the 30 μm-diameter microwires and the small rat spinal
cord, chronic implantation and stimulation produce limited
histological damage and do not compromise function.
Furthermore, despite the increased difficulties of implanting into
the smaller rat spinal cord, ISMS is effective in activating spinal
cord networks in the lumbosacral enlargement in a manner that is
safe, stable and reproducible.
Index Terms—spinal cord injury; intraspinal microstimulation;
functional electrical stimulation.
I. INTRODUCTION
S
pinal cord injury is a devastating neurological trauma, often
resulting in the paralysis of smooth and skeletal muscles
innervated by motoneuronal pools arising below the level
of the lesion. In addition to pharmacological and rehabilitative
interventions, various forms of electrical stimulation have been
suggested for reanimating paralyzed muscles [1]. One form of
Manuscript received August 4, 2015; revised January 6, 2016; accepted
March 17, 2016. Date of publication AA; date of current version BB. This
work was funded by the National Institutes of Health, Canadian Institutes of
Health Research, International Spinal Research Trust and Alberta Heritage
Foundation for Medical Research (AHFMR). VKM was an AHFMR Senior
Scholar.
J. A. Bamford was with the Centre for Neuroscience, University of Alberta,
Edmonton, AB, T6G 2E1, Canada (email: jeremy.bamford@gmail.com).
electrical stimulation, intraspinal microstimulation (ISMS),
applies low levels of electrical current into the lumbosacral
spinal cord via flexible microwires [2]. While invasive, this
central approach to electrical stimulation has certain
advantages: 1) the spinal cord contains, within a small region,
motoneurons which innervate muscles responsible for bladder
function, bowel function and locomotion [3]; 2) in addition to
motoneurons, this region contains the networks which
coordinate muscle activity and produce multi-joint synergies [4,
5]; 3) microwires are inserted far from the actuated muscles,
thus greatly reducing the possibility of damaging or dislodging
the microwires during movement; and 4) the spinal cord is
surrounded by the spinal column, affording both protection and
a solid structure to which implanted electronics may be affixed.
The ventral horn of the mammalian lumbosacral cord
contains organized motoneuronal pools which innervate the
muscles of the lower limbs [3, 6]. These pools have previously
been mapped in cats, humans and rats [7-10]. Furthermore, the
arrangement of motoneuronal pools is conserved across
species, making it relatively predictable where a given pool is
located.
Our work with ISMS has produced very promising results in
animal models. We have demonstrated that ISMS in various
regions of the lumbosacral cord can produce single muscle
activation, whole-limb flexion or extension synergies, rhythmic
locomotor-like activity and even weight-bearing, propulsive
over-ground walking. Examples of the movements elicited by
ISMS at various locations in the cat spinal cord can be found in
other works from our laboratory (e.g. [2, 11-15]
A hallmark of ISMS is its ability to produce force in a gradual
manner due to the near orderly, progressive recruitment of
motor units from small (slow) to large (fast) [16-18]. We have
shown that this is due to the trans-synaptic activation of
motoneurons through a distributed network of afferent
projections, propriospinal neurons and other interneurons
within the motoneuronal region of the ventral horn [2, 19].
Utilizing arrays of flexible microwires, we have shown that
R. M. Lebel was with the Department of Biomedical Engineering,
University of Alberta, Edmonton, AB, T6G 2S2, Canada (email:
rmlebel@ucalgary.ca).
K. Parseyan was with the Centre for Neuroscience, University of Alberta,
Edmonton, AB, T6G 2E1, Canada (email: kparseya@ualberta.ca).
V. K. Mushahwar is with the Department of Medicine and Centre for
Neuroscience, Edmonton, AB, T6G 2E1, Canada (email:
vivian.mushahwar@ualberta.ca).
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
2
ISMS can activate muscles selectively and synergistically to
produce functional movements [20, 21]. For example, ISMS
produces fatigue-resistant standing [14], in-place stepping [12]
and over-ground walking [15] in cats. Collectively, our
observations of ISMS suggest that it warrants further
investigation as a technique to restore function following a
spinal cord injury.
Epidural spinal cord stimulation is another method of
stimulating the spinal cord for restoring lost functions after
spinal cord injury. Herman and colleagues demonstrated that
epidural stimulation may improve walking capacity in
individuals with incomplete spinal cord injury who had
received treadmill locomotor training [22, 23]. More recently,
epidural stimulation applied in people with spinal cord injury
clinically classified as complete was able to boost residual
descending connections, such that some voluntary movements
of the legs could be performed when the stimulus is turned on
[24]. Nonetheless, due to the non-specific activation of
locomotor-related networks in the ventral horn of the spinal
cord of humans, and the lack of weight-bearing support
provided by the legs during rhythmic locomotor-like
movements induced by stimulation, the use of epidural
stimulation for restoring walking was deemed to be “an
extremely challenging long-term goal” by an NIBIB
Consortium for Addressing Paralysis through Spinal
Stimulation
Technologies
(http://myana.org/publications/news/national-instituebiomedical-imaging-and-bioengineering-report-spinal-cord).
The consortium encouraged that focus be instead diverted to the
use of epidural stimulation for restoring autonomic functions
after spinal cord injury.
In the current work we describe the fabrication, implantation
and chronic stability of ISMS microwires. We present our
methods for implantation in a feline model and our adaptations
to the murine (rat) model. Attention is given to difficulties we
encountered and how we overcame them. Given the interest
expressed in ISMS, we believe that the provision of a detailed
description of our methods would be widely beneficial. A
careful explanation of these techniques will encourage others to
investigate the basic science and clinical potential of this
promising paradigm of electrical stimulation. The rat model
opens several lines of investigation that are not as readily
facilitated in other models. For example, the rat model can be
used to address questions related to the effect of targeted
electrical stimulation on enhancing the action and guiding the
proliferation of neuroregenerative assays such as cell-based
therapies. It can also be used to assess the effect of targeted
electrical stimulation on the reorganization of neuronal
networks when used alone, or in conjunction with a
rehabilitation intervention, a pharmacological neuromodulator
and/or a genetic modifier (e.g., [25]).
II. METHODS
A. Procedures in the feline model
The procedures for chronic implantation of intraspinal
microwires in cats were originally derived from experiments
designed to obtain chronic recordings of single cells in the
dorsal root ganglia of awake, freely behaving cats [26]. Under
aseptic surgical conditions, a laminectomy is performed to
expose the lumbosacral enlargement. Microwires are routed
together through a silicone rubber tube, tunneled
subcutaneously to a headpiece and anchored to the skull. The
silicone tube is secured to the spinous process immediately
rostral to the laminectomy, and the intraspinal microwires
emerging from the tube are further secured as a bundle to the
dura mater using 8-0 sutures [21]. The array of 6-12
microwires, typically 30 μm in diameter, is then inserted one
microwire at a time with the microwire tip targeting the
appropriate ventral motoneuronal pools. Microwires are spaced
2-3 mm apart rostro-caudally and each is fixed in place with a
discrete drop of cyanoacrylate adhesive. Test stimulus trains are
delivered through the microwires to ensure that the tips are
properly positioned. Signs of accurate targeting include
selective activation of a target muscle or multi-joint synergy,
and a gradual increase in recruited force and movement
development as stimulus amplitude is increased [18].
B. Identifying locations of intraspinal microwires in cat using
MRI
Microwire locations in some cats were determined using
magnetic resonance imaging (MRI). The lumbosacral region of
the spinal cords of animals were imaged in vivo approximately
one month prior to microwire implantation and at 0.5, 2 and 4
months following implantation. All imaging was performed in
a 4.7 Tesla Varian Unity Inova spectrometer with a 14 cm
diameter birdcage radiofrequency coil. Maximal gradient
strength was 60 mT/m with 120 mT/m/ms slew rate. Two
different spoiled gradient echo MRI protocols were used: a
moderately T2*-weighted multislice 2D sequence for presurgical assessment; a low tip angle mildly T2*-weighted 3D
sequence for post-surgical imaging. The pre-surgical sequence
was designed empirically for optimal contrast between
cerebrospinal fluid, grey matter, and white matter. The
repetition time was 2.1 s, the echo time was 14.0 ms, the
excitation angle was 60°, the readout bandwidth was 50 kHz
(97.7 Hz/pixel), and the field-of-view was 128.0 mm x 153.6
mm with an acquisition matrix of 512 pixels x 512 pixels. The
animal was positioned in a supine posture inside the bridge coil
and phase encoding was performed in the left/right direction to
reduce breathing artifacts from projecting through the spinal
cord. Forty contiguous 1 mm thick slices were collected in an
interleaved acquisition ordering and fat saturation was
employed. Six identical scans were collected consecutively for
a total scan time of 108 minutes; scans were co-registered to
reduce inter-scan motion, averaged to improve signal- and
contrast-to-noise and subsequently interpolated in-plane by a
factor of 2 for a voxel size of 0.125 x 0.150 x 1.000 mm3.
The post-surgical imaging protocol was designed with
primary consideration given to implant-magnetic field
interactions and secondary consideration to tissue contrast. The
potential for localized heating proximal to the microwires, a
result of coupling between the radiofrequency coil and the
microwire leads, precluded the use of high tip angle pulses (as
required for either spin echo or 2D gradient echo approaches).
Furthermore, field shifts induced by magnetic susceptibility
created substantial regions of signal void around the microwire
and limited the echo time and T2*-weighting. Therefore, we
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
instead performed 3D spoiled gradient echo imaging with a
repetition time of 14.7 ms, an echo time of 5.74 ms, and a 5°
excitation angle. These parameters were selected empirically to
provide tolerable signal loss surrounding the microwires and
minimal potential for tissue heating, yet sufficient contrast
between grey and white matter. We used a readout bandwidth
of 80 kHz (156.3 Hz/pixel) and a field-of-view of 128 mm x
128 mm x 80 mm with a matrix size of 512 pixels x 512 pixels
x 80 pixels. Phase encode directions were left/right and
rostral/caudal, and fat saturation was not required since fat
surrounding the region of interest of the spinal cord was
removed during surgery. Twelve identical scans were collected
consecutively for a total scan time of 120 minutes. Images were
co-registered, averaged, and interpolated to a voxel size of
0.125 x 0.125 x 1.000 mm3.
C. Microwire fabrication for use in rats
What follows is a more detailed explanation of fabrication,
implantation and stability of intraspinal microwires with respect
to rat experiments. Microwires were fabricated from 30 μmdiameter platinum/iridium (80%/20%) wire insulated with a 4
μm layer of polyimide (California Fine Wire, Grover Beach,
CA, USA). Microwires fabricated from 25 μm wire are more
flexible and preferable for use in rats. The 25 µm diameter
microwire can withstand the stimulus amplitudes needed to
elicit functional movements through ISMS, and retains
adequate stiffness to be inserted vertically into the cord such
that it reaches the intended region of interest. Nonetheless, the
smaller diameter wires are considerably more delicate and can
be more easily damaged during the construction process of the
array. We encourage that the techniques are first practiced and
perfected on the 30 µm diameter wire prior to transfer to the
smaller diameter one.
Microwire tips were de-insulated by mechanically rubbing
the polyimide layer with a fine micrometer or between two
glass slides 1 mm from the tip. When carefully performed, both
methods separated the insulation from the microwire tip
without flattening the platinum-iridium core. Tips were then
sharpened by cutting at a 15° angle with a #15 hard stainless
steel surgical blade, leaving 30-60 μm of metal de-insulated
from the tip (Fig. 1). The tip ends of the microwires were then
gently bent to a 90° angle according to the desired length
(length 1 in Fig. 1B). A custom-built jig that allowed the user
to align the tip perpendicularly to an edge of an acute angle was
used for bending the wires close to the tip end. A ruler with 100
µm tick marks allowed the user to choose the desired length
from the tip prior to bending. Bending was performed by gently
pressing the experimenter’s finger on the electrode against the
acute end. The angle was verified visually and adjusted by
gently holding the electrode at the bent region with a pair of
forceps and increasing or decreasing the angle until it reached
90°.
Microwires were arranged in arrays as shown in Fig. 1. The
tips were temporarily inserted into Silastic brand laboratory
tubing (0.94 mm and 0.51 mm inside and outside diameters,
respectively) both for protection prior to implantation and for
securing the microwires in place with the desired inter-wire
spacing during the array fabrication process. Typically,
microwires were spaced in 2 (rats) or 3 mm (cat) intervals
3
(length 2 in Fig. 1) along each side of the array. The length of
the full array was preset and a second 90° bend upwards (as
indicated by arrow in Fig. 1B) was made in order to
accommodate the routing of the wires up the base of the nearest
rostral spinous process. This distance was typically 6-8 mm in
rat and 10-15 mm in cat (length 3 in Fig. 1). This second bend
ensured that the epidural (lead) portion of the microwires lay
flat and flush with the surface of the dura mater. The bend was
performed by holding the bundle of mircowires with a pair of
forceps where the teeth of the forceps pressed gently over a
temporary silk suture placed to hold the microwires together,
and bending the bundle against the experimenter’s finger. This
process retained microwire alignment and ensured that the
insulation around the microwires remained intact. The second
bend can cause additional stress on the wires if not properly
aligned in the sagittal and transverse planes with the first bend
(close to the tip). Therefore, great care was taken in placing both
bends to ensure appropriate alignment.
A final 90° bend was made so that the microwires were routed
into a silicone tube which lay on the dorsal laminar surface of
the rostral spinous process. This end of the tube was filled with
electronics grade silicone to ensure that no fluids could enter
the tube. Typically, at least 2 arrays of varying lengths (length
3 in Fig. 1) were prepared before each surgery to allow the
surgeon some flexibility in the placement of the array.
To connect to the microwires in the array, the microwires
were soldered to individual female pins which were inserted
into a plastic base that formed the headpiece (PlasticsOne,
Roanoke, VA, USA). A return electrode was fashioned from
multi-strand stainless steel wire (Cooner AS632, Chatsworth,
CA) and the other end soldered to one of the connector pins.
The headpiece was encapsulated in silicone and epoxy to ensure
stability and electrical isolation of the pins.
Fig. 1. Schematic of the intraspinal implant. A) An example of a 4microwire array designed for chronic implantation in the rat. B) The
microwires are routed from a silicone rubber tube, and pre-made to fit the
experimental animal. Typical measurements of: 1) microwire depth; 2)
inter-wire spacing; and 3) distance to first wire are given for cat and rat
(arrow indicates the location of the suture placed to anchor the microwire
bundle to the dura mater). Implants are fixed to the T12 spinous process in
rat (shown) and the L3 or L4 process in the cat.
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
D. Surgical procedure for microwire implantation in rats
Rats were anesthetized using isofluorane (inhalant, 1.5%2.0%) and prepared for aseptic surgery. The headpiece and
silicone tube containing the microwires were cold sterilized by
soaking in the strong disinfectant solution, Germex (Vétoquinol
N.-A. Inc., Quebec, Canada), for a minimum of 20 minutes.
They were then rinsed in sterile saline. Germex was commonly
used to disinfect fine instruments in animal surgical procedures.
We attempted the use of ethylene oxide (EtO) and found that it
changed the properties of the silicone rubber used for the head
piece and inside the silicone tube, making it more brittle. We
therefore stopped using that process for sterilization. In future
human experiments, medical grade rubber materials that can
withstand standard sterilization procedures, such as EtO will be
used. Autoclaving was also attempted but found to destroy the
plastic headpiece used in these experiments. Germex was very
effective and no infections were encountered with respect to the
implanted portions of the ISMS system.
The silicone tube was tunneled subcutaneously to the head
where the headpiece was affixed to the skull using stainless
steel machine screws (0-80 x 2.4 mm, PlasticsOne, Roanoke,
VA, USA) and dental acrylic. The skin was sutured closed
around the acrylic headpiece. Despite the simplicity of the
headpiece construction, no infections of this area were
observed. A laminectomy was then performed to expose the
lumbosacral enlargement, corresponding to spinal segment L2
to L5 in rat. This typically involved the removal of most of the
T13 to L1 spinous processes. The silicone rubber tube
containing the bundle of microwires was temporarily affixed to
the laminar surface of the T12 vertebra with a drop of
cyanoacrylate adhesive. A hole was drilled into the spinous
process of the rostral vertebral process, through which a small
screw or suture needle was inserted. This acted as the anchor
for the application of dental acrylic which was applied to
encapsulate the silicone tube and fix it in place.
Prior to the implant procedure, a test microwire was inserted
into the spinal cord in order to verify the motoneuronal pool
arrangement. Stimulus trains delivered through the test
microwire were expected to produce selective activation of
target muscles and to recruit force in a gradual manner. The
surgeons verified this through palpation of the muscles of the
lower limb, observing activation of the appropriate muscles and
a gradual increase in force as current amplitude was increased
from stimulus threshold level using successive test stimulus
trains. In rat experiments we targeted the motoneuronal pool
innervating the quadriceps muscle group.
Once the correct rostro-caudal placement was verified, the
ISMS array was implanted by inserting one microwire at a time
(see Fig. 2 for an example in cat). Microwire insertion
commenced by holding the wire with fine forceps close to the
tip. Once the tip penetrated the superficial dorsal layers of the
cord, the remaining portion was inserted vertically until the 90°
bend was reached. In rats, as well as cats, the dura mater was
not opened during microwire implantation due to the very small
subdural space in these species (<100 µm in rats and <250 µm
in cats). Instead, a small hole was made in the dura mater using
the tip of a 30 gauge hypodermic needle.
In the event that a microwire tip does not reach the right
4
location, the intended movement(s) is not elicited with
stimulation, is elicited at higher than usual threshold levels, or
is elicited in conjunction with functionally undesirable cocontractions. The response usually indicated the direction of
error and a subsequent reinsertion generally corrected the error.
Multiple penetrations in one location of the cord increase the
risk of tissue damage; therefore, we have limited the number of
reinsertions to a maximum of two for any given location.
Fig. 2. Routing of microwires along the spinal cord. ISMS microwires
are routed along the surface of the cord to the nearest rostral vertebra.
Microwires are implanted individually so that each wire can float on top
of the cord, independently of other microwires. A) An example of
microwires implanted in one cat spinal cord is shown just after
implantation. B) A coronal MRI slice taken 14 days following
implantation in another cat. The black spots in the cord are the signal voids
created by the microwires. These appear enlarged because of MRI signal
loss surrounding the 30 μm-diameter microwires; the microwires are ~17
times smaller than the void spots.
The microwire tip depths in Fig. 1 were preset to correspond
to the depth of Rexed’s Lamina IX in the ventral horn of the
animal, such that the microwires laid flat on the epidural surface
of the spinal cord. Maps of the motoneuronal pools in the rat
spinal cord [9] were used to approximate the distances required
for implant fabrication. Individual microwires were affixed to
the epidural surface at their point of insertion with discrete
drops of cyanoacrylate adhesive and run along the epidural
surface of the cord to the base of the nearest rostral spinous
process (Fig. 1A). The wires converged at this point and were
sutured to the dura mater using 8-0 suture (indicated by arrow
in Fig. 1B). Securing the microwires to the dura mater was
critical as this ensured that each microwire could rest on the
spinal cord, and that microwires would not be pulled away from
the epidural surface by any connective tissue that may have
formed in the laminectomy site during recovery. Finally, a layer
of thin plastic film made of low density polyethylene, cold
sterilized in Germex, was measured to the size of the site of the
exposed laminectomy and used to cover the implanted region
of the spinal cord. The film was affixed to the vertebral edges
(sides of the spinal cord in cats) around the site with drops of
cyanoacrylate adhesive. This helped to prevent the invasion of
connective tissue into the laminectomy site during recovery.
Muscle layers were separated and individually sutured over the
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
wound before suturing the skin incision closed. Following the
surgical implantation, rats were allowed to recover for 7 days
before any further experiments and were administered
analgesics and antibiotics.
E. Functional tests
After the recovery period, test stimulus trains were applied in
order to ascertain the functional responses from the microwires.
Similar to the procedure performed during the microwire
implantation, the experimenter observed the evoked
contractions to ensure that the targeted muscles were activated
and that they recruited force in a gradual manner. These
responses gave an ongoing functional indication of the location
and stability of microwires. Monopolar stimulation of charge
balanced, 200 μs biphasic pulses, was delivered in 1 s trains at
a rate of 25 pulses per second (pps) through the microwires in
one side of the spinal cord. The stimuli were interleaved
between two adjacent microwires resulting in a total stimulus
frequency of 50 pps. Stimulus amplitudes ranged from 100-250
μA. The level of total injected charge was limited in order to
avoid causing tissue damage [27]. Stimulus thresholds, the
lowest stimulus amplitude producing a palpable muscle action,
were noted routinely as an indicator of microwire stability.
To assess the effect of long-term ISMS on spinal cord tissue,
stimulation was delivered for 4 hours every day for 30
consecutive days in rats. During terminal experiments,
isometric force levels in response to increasing stimulus
amplitude were measured and compared to that derived from
experiments in which the microwires were implanted acutely.
The rats were positioned in a stereotaxic array and each patellar
ligament was dissected from its point of insertion and attached
to a force transducer (Interface MB-5, Interface Inc., Scottsdale,
AZ, USA). The force recordings were amplified 100-fold and
captured at a sampling rate of 1000 Hz using a data acquisition
interface (Power 1401, Cambridge Electronic Design,
Cambridge, UK) with associated software (Signal v. 2.13,
Cambridge Electronic Design). Force recruitment curves were
generated using 5 microwires in 3 chronically implanted rats
and 6 microwires in 3 acutely implanted rats. Trains of
stimulation were delivered through a single wire at a time (i.e.,
no interleaving). Stimulation amplitude was varied upwards
from sub-threshold levels to a maximum of 400 μA in random
order and the force evoked by 50 pps stimulus trains was
recorded. Evoked tetanic forces were normalized to peak force
and plotted against stimulus amplitude.
F. Post-mortem Identification of electrode tip locations in the
spinal cord
In both cats and rats, electrode tip locations were generally
identified by removing thin cross-sections of tissue in tissue
blocks containing one or two electrodes until the tip of the
electrode could be seen. Pictures of the blocks were then taken
and measurements made to identify the angle of insertion of the
electrode and the location of the tip in the ventral horn. In some
cats, MRIs were also obtained of the extracted spinal cords with
the ISMS microwires in place prior to sectioning the tissue.
Because of the larger size of the cat spinal cord, the ISMS
electrodes could not be easily dislodged during the sectioning.
5
This was not the case in rats, where more incidents of
dislodging took place during the process of identifying tip
locations. In chronically implanted rats in which detailed
histological and immunohistochemical analyses were obtained
[28], the location of the tips was identified by markers of
immune reactivity such as GFAP and ED-1.
III. RESULTS
A. Targeting accuracy
The placement of ISMS microwires in rat experiments was
relatively successful. Due to the smaller diameter, microwires
were considered to be accurately placed if the tips were located
in the ventral grey matter. This area encompasses lamina VIII,
IX and much of lamina VII. Out of 33 microwires implanted in
an acute experiment, 17 were located in post-mortem
histological examination using microdissection [17]. All 17 of
these acutely implanted wires were located in the ventral grey
matter of the lumbar spinal cord (Fig. 3). Likewise, the
placement of microwires in chronic rat experiments was
generally successful. Of 24 chronically implanted microwires,
16 were located post-mortem using markers of inflammation
along the microwire tracks such as ED-1 or GFAP
immunoreactivity as guidance [28]. Similar to the acute
experiments, the microwire tips that were identified were all
located in the ventral grey matter (Fig. 3).
Fig. 3. Summary of locations of ISMS microwires in rat. The locations
of ISMS microwires recovered from multiple experiments in rat. Gray
circles indicate the positions of 17 ISMS microwire tips from acute rat
experiments. The positions of 16 microwire tips implanted for 38 days are
also shown (triangles and diamonds). Of these tips, 7 were used to deliver
daily stimulation (white triangles) while 9 were used as sham implanted
controls (black diamonds).
B. Microwire stability assessed functionally and directly
through MRI
Measurements of function in response to ISMS can be used
to indicate the stability of implanted microwires over time.
Changes in stimulus thresholds for muscle activation, or in the
relationship between evoked forces and stimulus amplitude can
indicate damage to surrounding neural structures, degradation
of the microwire, or a shift in its location due to mechanical
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
dislodgement or glial encapsulation. In rats stimulated daily for
30 days, mean threshold for muscle activation decreased from
118.2 μA ± 45.1 μA to 67.8 μA ± 25.1 μA (mean ± standard
deviation). A correlation between time and threshold for muscle
activation was detected over this period (Fig. 4, p = 0.002).
In the acutely implanted rat and cat, a graded relationship
between evoked force and increasing stimulus amplitude has
been demonstrated in previous studies [17, 18]. In the current
work we compared the forces evoked in response to ISMS in
acute rats, and after chronic stimulation through intraspinal
microwires in the same species. We found that a graded
relationship between force and stimulus amplitude was
maintained in the chronically stimulated rat, suggesting no
shifting of the microwire tips within the spinal cord or damage
to the neuronal networks which produce this gradual forceamplitude relationship (Fig. 5). Analysis of the force
recruitment relationship in acute and chronically stimulated rats
indicated no significant difference in the slope of the regression
between these groups.
6
Fig. 5. Recruitment of force by varying amplitude above activation
threshold. Force recruitment curves obtained by stimulating through
microwires implanted acutely and chronically in separate groups of 9 and 5
rats, respectively. Pulse trains at 25 pps were delivered through the
intraspinal microwires and the isometric tension generated was measured
with a force transducer attached to the quadriceps tendon. Forces evoked
from each microwire were normalized to their respective maximum forces.
Mean peak force recruited during terminal experiments was 0.96 ± 0.35 N,
and 1.64 ± 0.39 N for the acutely implanted and chronically implanted rats,
respectively. This difference in recruited force is likely explained by the
fact that the chronically implanted animals received 4 hours of stimulation
per day for 30 days prior to the terminal experiments. As we have previously
observed, this daily training produced plastic adaptations in the muscle that
may explain this result (Bamford et al., 2010). Most importantly, the slopes
of the force-stimulus amplitude relationship produced by both acutely and
chronically implanted microwires are similar, indicating the stability of
IV. DISCUSSION
A. Overview
Fig. 4. Reduction in stimulus thresholds for quadriceps muscle
activation over time. The lowest stimulus amplitude producing a
discernible quadriceps muscle twitch (stimulus threshold) decreased over
time in chronically implanted rats. Following a 7 day recovery period after
microwire implantation in the lumbar spinal cord, stimulus thresholds of
individual microwires were monitored. Quadriceps muscles of awake,
unrestrained animals were palpated in response to recruitment by 25 pps
stimulus trains. Activity-induced plasticity may be responsible for this drop
in threshold over time. (p = 0.002, R2 = 0.243). The results are from 12
microwires implanted in 6 rats for 38 days. The threshold level was
obtained for each microwire on the first day of stimulation (day 7), and at
2 to 3 time points over the ensuing 30 days. Some data points on the graph
represent measurements from more than one microwire.
The stability of microwires can also be confirmed directly by
monitoring their locations over time in the same animal. Serial
MRI of a chronically implanted cat is shown in Fig. 6,
demonstrating the stability of microwire placement over a 4
month period. The positions of both the accurately targeted and
the mis-targeted microwires are shown. No detectable
displacement was observed in any of the implanted microwires
in this animal, suggesting a high degree of stability regardless
of location in the spinal cord. Due to the small size of the rat
spinal cord, MRI was not used as a means of identifying or
monitoring the location of the implanted microwires.
Intraspinal microstimulation is a paradigm of electrical
stimulation which applies electrical current in deep spinal
structures to restore function after neurological trauma, such as
a spinal cord injury. It also allows for novel investigations of
neural networks and augmentation of neuro-regenerative
interventions. A thorough description of the fabrication,
implantation and stability of the system will allow others to
apply it more readily in their research programs.
Currently, the fabrication of the ISMS array is performed
manually, requiring skill in fine manipulations and generally
requiring 3-5 days to manufacture. Materials such as fine wires
are chosen for their flexibility and small diameter as this
minimizes damage to the spinal tissues. This, however, creates
a fragile implant which must be protected during the fabrication
and implantation procedures. Work is currently underway to
automate and standardize the construction of microwire arrays
for both rats and cats, with the aim of shortening manufacturing
time while maintaining the ability to build custom arrays, sized
specifically for each animal. Some of the difficulties we have
encountered are discussed below with special reference to the
rat model.
B. Challenges in fabrication and implantation
Proper microwire fabrication minimizes difficulties during
the implantation procedure. Microwire tips that are not cleanly
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
7
The laminectomy is a fairly routine procedure in the cat and
larger animals; however, the rat model affords much less space
and the risk of damage to the surface of the cord while removing
the spinous processes and dorsal laminae is high. When
implanting, care must be taken to ensure that microwires are
free from any torque that might cause them to move from their
implanted position. Any moment on the microwire, whether
causing rotation or extension, can cause the microwire to move
through the spinal cord tissue after implantation. This point
reinforces the need for care during fabrication of the ISMS
array, as adjustment during surgery is difficult. Movement of
the microwire through the spinal cord after implantation will
likely result in tissue damage and the microwire tips missing the
intended target.
Cyanoacrylate adhesive must be carefully applied in discrete
drops so that each microwire is secured, yet kept independent
of the other microwires. This allows the microwires to float
independently from each other upon the surface of the cord. If
the adhesive is applied carelessly, the wires may become glued
to one another, increasing their chance of pulling out from the
spinal cord. A small 1-ml syringe fitted with a 30-gauge needle
with the cutting tip cleanly removed can be used to apply very
small drops of adhesive. The suturing of the microwire bundle
to the dura mater is a delicate step, especially considering the
small subdural space in the rat; however, we have found that
skipping this step inevitably results in the wires pulling free
from the cord.
Fig. 6. Stability of implanted microwires over time. MR images of an
accurately targeted microwire (B - D) and a mistargeted microwire (F - H)
in the same cat are shown. Images were taken 1 month before implantation
(A,E), 0.5 months after implantation (B,F), 2 months after implantation
(C,G) and 4 months after implantation (D,H). Regardless of targeting
success, microwires remain extremely stable after the first week of
implantation.
de-insulated or finely sharpened will impede insertion through
the dura mater in the rat and cause excess damage within the
cord (Fig. 7, A-D). To address this issue, we are investigating
various techniques for automation of the microwire fabrication.
For example, rather than cutting the microwire tips with a
scalpel blade to produce a sharp bevel, it is possible to insert the
de-insulated microwires through a pulled glass capillary tube
that is clamped onto a micropipette beveller (BV-10E, Sutter
Instrument, Novato, CA, USA). Through this approach,
beveled tips, among other possible tip shapes, can be produced
reliably at any angle (Fig. 7, E). If the microwire is not
accurately bent to 90°, or if the final length is incorrect, the
microwire tip will not be accurately placed in lamina IX,
leading to uncertain and possibly inappropriate stimulation
outcomes. Furthermore, the bending process must be performed
gently as rough handling of the microwire at this stage can
cause breaks in the insulation at the bend point, leading to
current leakage and, essentially, epidural stimulation through
this break. In our experience, a microwire with a break at this
point generates flexor withdrawal movements instead of the
intended movement. Finally, the construction of the array must
ensure that all microwire tips are aligned parallel to one another,
pointing downwards, and that no torque remains that might
push the microwire out of position after insertion.
Fig. 7. Microwire sharpening and de-insulation examples and
challenges. A) An example of a properly sharpened and de-insulated
microwire. B) A microwire with ‘poking’ insulation that would drag across
tissue during insertion. C) A microwire with a ‘hooked’ tip and loose
insulation. Such a tip will fold upon insertion and prevent safe extraction
of the electrode if multiple insertions are necessary. The loose insulation
would increase insertion resistance, dragging and damaging tissue as force
is applied. D) A ‘flag’-tipped microwire, typically caused by cutting the
microwire with a dull scalpel blade. Insertion with this tip shape can induce
damage, especially if multiple penetrations are required. E) A machinebeveled electrode. One step in the automation of the microelectrode
fabrication involves machine-beveling microelectrodes by inserting them
through a pulled glass micropipette clamped onto a micropipette beveler
(BV-10E, Sutter Instrument, Novato, CA, USA). All microwires are 30 µm
in diameter with a 4 µm coat of insulation.
C. Targeting success of microwire tips
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
In our experience, the best placement for microwire tips for
evoking functional limb movements is in the ventral grey
matter, especially lamina IX. Achieving consistent dorsoventral depth and medio-lateral placement of microwire tips in
this region has been the focus of much work in our lab. The
accuracy we have achieved in dorso-ventral targeting has varied
in different experiments. In one investigation using chronically
implanted cats, 25% of implanted microwires were found in the
ventral white matter tracts during post-mortem examination
[21]. Functionally, 80% ± 12.5% of the implanted microwires
were effective in producing consistent, activation of a specific
muscle group throughout the 6-month period of microwire
implantation, indicating accurate targeting of the microwire
tips. In another ISMS study utilizing cats, the success rate in
achieving accurate placement of the microwire tips was
approximately 50% in acute cat implants as judged by postmortem histology [12]. Many of the misplaced microwire tips
were located in the ventral white matter tracts. In light of the
importance of microwire depth to targeting accuracy we
employed cylindrical microelectrodes with multiple contact
points set at 4 depths [18, 29]. The multiple depths this
configuration affords eliminated mistakes in dorso-ventral
targeting caused by inappropriate microwire pre-fabrication
depth. We have also introduced high-resolution magnetic
resonance imaging pre-operatively to obtain detailed
dimensions of the spinal cord in cats (e.g., Fig. 6, A, E). This
has allowed for the microwire arrays to be fashioned after the
specific dimensions of each animal as opposed to average
dimensions utilized in earlier studies.
Microwire tips may also be misplaced medio-laterally. For
example, some microwires in cats deviated from a vertical
pathway and ended with their tips in the medial or lateral
columns (see Fig. 6, F-H for one example). The correct mediolateral placement can be misjudged due to optical illusions
introduced by the surgical microscope. Such illusions
compromise judgment of vertical positioning of the microwires
prior to insertion in the cord (Fig. 6, F-H). We have found that
a second observer (or camera) can aid the surgeon who is using
the surgical microscope to place the microwire. This twoperspective method has improved judgement of the vertical
orientation of the microwire during insertion. We have also
recently introduced the use of a microwire insertion guide to
further ensure the vertical placement of the electrode in the
spinal cord. Initial testing of the guide using surrogate spinal
cords [30] demonstrated that its utility can substantially reduce
off-vertical deviations and positioning the microwire tip within
the desired 0.25 mm tolerance for a region of interest (Fig. 8).
The targeting accuracy of electrodes implanted in the rat may
be less stringent than that of the cat depending on the research
question asked. For example, in a study where our interest was
primarily in evaluating the stability of ISMS implants in the
spinal cord in chronic experiments and the extent of damage
they may cause to surrounding tissue [28], the region of interest
was defined as the ventral gray matter as opposed to a particular
motoneuronal pool, thus increasing the targeting tolerance from
0.25 in the cat to 0.50 mm. With these specifications, targeting
accuracy of microwire tips was deemed greater in the rat than
that for microwires implanted in the cat. To avoid off-vertical
deviations in the rat, an insertion guide could be used as well,
therefore allowing for more specific regional targeting. In
8
general, accurate targeting in the cat without the use of a
microwire insertion guide can be more challenging than in the
rat. This is because the targeted depth in the rat spinal cord is
approximately 1.5 mm vs. as much as 4.5 mm in the cat spinal
cord (length 1 in Fig. 1). The 3-fold greater distance of insertion
requires a more accurate angle of microwire penetration in the
spinal cord of the cat, as any deviation from the accurate course
will be magnified over the greater depth of insertion (e.g., Fig.
8, top).
Insertion without guide
Insertion with guide
p < 0.005
Fig. 8. Quantification of microwire deviation when inserted by hand and
by using an insertion guide. Testing was performed using surrogate spinal
cords (top, n=13) with dimensions and mechanical properties similar to
those of the cat spinal cord (prepared as described in Cheng et al., 2013).
The use of the microwire insertion guide significantly reduces microwire
deviation (bottom, n=14) and increases accuracy of positioning of the
microwire tip. Error bars represent standard deviation.
D. Stability
In addition to targeting the microwires accurately, it is
important for the microwires to remain in place if stable
responses are to be evoked over time. As demonstrated by MRI,
microwires do not appear to shift position once they are
securely implanted (Fig. 6). In our experience, microwires can
shift or, more likely, pull out of the cord to some degree during
the first 12 hours post-surgery. This effect can be substantially
reduced by maintaining the animals in a lightly sedated state for
at least 12 hours post-implantation.
One indication of stability is the maintenance of functional
responses evoked by ISMS over time. In the chronically
implanted rat, activation thresholds decreased over 30 days
(Fig. 4). The most probable explanation for this finding is that
the rat experiments involved daily stimulation for 4 hours. It
may be that daily stimulation, within safe levels, encourages
plasticity of the spinal networks leading to lowered activation
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
thresholds. The mechanism for this decrease over time might
involve a decrease in edema, or a decrease in inhibitory
neurotransmission in the spinal cord similar to that observed
following daily stepping training in chronically transected cats
and rats [31]. Another possible scenario involves the migration
of neurons towards the stimulation site and the growth of neural
processes in response to the electrical stimulation [32, 33]. As
neurons migrate or extend towards the stimulation site, the
radius of stimulation necessary to elicit a threshold response
decreases, allowing a reduced level of current to produce the
same effect.
One of the functional hallmarks of ISMS is the ability to
recruit force in a gradual manner with increases in stimulus
intensity [17, 18]. Alterations to this relationship between force
and stimulus intensity could indicate movement of, or damage
to the microwire; or damage to the neural structures around the
microwire. We found that the force recruitment characteristics
of ISMS in rat were not changed by chronic implantation and
stimulation, indicating functional stability over time.
We have previously assessed markers of the immune
response surrounding chronically implanted microwires. An
extended immune response, commonly referred to as a
‘frustrated response,’ to implanted devices is common when the
implanted foreign body cannot be absorbed by the organism;
however, this frustrated response often subsides over time [34].
A survey of tissues surrounding microwires implanted for 38
days in rat spinal cord indicated limited damage and
inflammation [28]. A ‘frustrated response’ was not detected in
microwires implanted in cat spinal cords for 6 months [1, 35].
These studies suggest that the immune response we detected in
rats would eventually subside as we noted in the 6-month cat
experiments; however, it is also possible that these results
reflect a species difference between rats and cats in the immune
response to spinally implanted microwires. Moreover, the
higher mechanical mismatch between the 30 µm wires and rat
spinal cord could possibly lead to a larger immune response
than in the cat. We conclude that, with proper care in
manufacturing and implantation, ISMS microwires are stable
over long periods (6 months, longest tested), both in terms of
location and functional responses.
E. General specifications of an ISMS implant for human use
The evidence gathered to date suggests that ISMS may be a
viable approach for restoring standing and walking after spinal
cord injury. The lumbosacral enlargement in humans is 5 cm
long and the target regions in the ventral horn are 4.5 – 6 mm
deep (relative to the dorsal surface). These dimensions are 33%
larger than those in the cat; therefore, the general overall
structure of the ISMS implant used to date will be maintained
and procedures utilized to date will be followed These include
pre-surgical MRI and the use of insertion guides to ensure that
the stimulation sites target the regions of interest with 0.25 mm
tolerance. Nonetheless, the implant would require a number of
modifications from that used in the cat for eventual use in
humans. The modifications include the use of multi-site
electrodes, electrode leads that can accommodate the nearly
12% strain experienced by the spinal during extreme flexor and
extensor movements of the spinal column and hips, and
intradural fixation since the implant would be placed
9
subdurally. The electrodes will be connected to a wirelessly
operated implanted multi-channel stimulator with a minimum
of 16 channels (preferably 32 channels). All implantable
components would withstand the necessary sterilization
procedures to prevent infections. Of importance would also be
the control strategies modulating the stimuli through various
electrodes to produce safe and functional standing and walking.
Strategies already developed in our group will form the basis of
such control paradigms in the future [36-38].
V. CONCLUSION
Intraspinal microstimulation is a viable method for directly
exciting the spinal cord networks in the lumbosacral cord.
These networks are responsible for control of the lower limbs
including standing and walking. We have expanded our
previous work in the cat; in the process developing techniques
that allow us to implant microwires chronically into the more
challenging rat model. We have found this process to be
feasible, safe and reproducible despite the increased mechanical
mismatch in size and stiffness between the rat spinal cord and
the 30 μm diameter microwires that have previously been
implanted in the cat.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
A. Prochazka, V. K. Mushahwar, and D. B. McCreery, "Neural
prostheses," Journal of Physiology, vol. 533, pp. 99-109, 2001.
V. K. Mushahwar, P. L. Jacobs, R. A. Normann, R. J. Triolo, and N.
Kleitman, "New functional electrical stimulation approaches to
standing and walking," J Neural Eng, vol. 4, pp. S181-97, Sep 2007.
E. Henneman and L. M. Mendell, "Functional organization of
motoneuron pool and its inputs," in Handbook of Physiology. Section
1: The nervous system. vol. II. Motor Control, Part 1, V. B. Brooks,
Ed., ed Bethesda, MD: American Physiological Society, 1981, pp.
423-507.
E. Jankowska, "Interneuronal relay in spinal pathways from
proprioceptors," Progress in Neurobiology, vol. 38, pp. 335-78, 1992.
O. Kiehn, "Locomotor circuits in the mammalian spinal cord," Annu
Rev Neurosci, vol. 29, pp. 279-306., 2006.
G. J. Romanes, "The motor pools of the spinal cord," Progress in
Brain Research, vol. 11, pp. 93-119, 1964.
G. J. Romanes, "The motor cell columns of the lumbo-sacral spinal
cord of the cat," Journal of Comparative Neurology, vol. 94, pp. 313358, 1951.
W. J. W. Sharrard, "The distribution of the permanent paralysis in the
lower limb in poliomyelitis: a clinical and pathological study,"
Journal of Bone and Joint Surgery, vol. 37B, pp. 540-558, 1955.
S. Nicolopoulos-Stournaras and J. F. Iles, "Motor neuron columns in
the lumbar spinal cord of the rat," J Comp Neurol., vol. 217, pp. 7585., Jun 10 1983.
V. G. Vanderhorst and G. Holstege, "Organization of lumbosacral
motoneuronal cell groups innervating hindlimb, pelvic floor, and
axial muscles in the cat," J Comp Neurol, vol. 382, pp. 46-76, May
26 1997.
V. K. Mushahwar, D. M. Gillard, M. J. Gauthier, and A. Prochazka,
"Intraspinal micro stimulation generates locomotor-like and
feedback-controlled movements," IEEE Transactions on Neural
Systems & Rehabilitation Engineering, vol. 10, pp. 68-81, 2002.
R. Saigal, C. Renzi, and V. K. Mushahwar, "Intraspinal
microstimulation generates functional movements after spinal-cord
injury," IEEE Transactions on Neural Systems & Rehabilitation
Engineering, vol. 12, pp. 430-40, 2004.
L. Guevremont, C. G. Renzi, J. A. Norton, J. Kowalczewski, R.
Saigal, and V. K. Mushahwar, "Locomotor-related networks in the
lumbosacral enlargement of the adult spinal cat: activation through
intraspinal microstimulation," IEEE Trans Neural Syst Rehabil Eng,
vol. 14, pp. 266-72, Sep 2006.
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
B. Lau, L. Guevremont, and V. K. Mushahwar, "Strategies for
generating prolonged functional standing using intramuscular
stimulation or intraspinal microstimulation," IEEE Trans Neural Syst
Rehabil Eng, vol. 15, pp. 273-85, Jun 2007.
B. J. Holinski, K. A. Mazurek, D. G. Everaert, R. B. Stein, and V. K.
Mushahwar, "Restoring stepping after spinal cord injury using
intraspinal microstimulation and novel control strategies," Conf Proc
IEEE Eng Med Biol Soc, vol. 2011:5798-801., p.
10.1109/IEMBS.2011.6091435., 2011.
V. K. Mushahwar and K. W. Horch, "Muscle recruitment through
electrical stimulation of the lumbo-sacral spinal cord," IEEE
Transactions on Rehabilitation Engineering, vol. 8, pp. 22-29, 2000.
J. A. Bamford, C. T. Putman, and V. K. Mushahwar, "Intraspinal
microstimulation preferentially recruits fatigue-resistant muscle
fibres and generates gradual force in rat," Journal of Physiology, vol.
569, pp. 873-84, 2005.
S. Snow, K. W. Horch, and V. K. Mushahwar, "Intraspinal
microstimulation using cylindrical multielectrodes," IEEE
Transactions on Biomedical Engineering, vol. 53, pp. 311-9, 2006.
R. A. Gaunt, A. Prochazka, V. K. Mushahwar, L. Guevremont, and
P. H. Ellaway, "Intraspinal microstimulation excites multisegmental
sensory afferents at lower stimulus levels than local alphamotoneuron responses," J Neurophysiol., vol. 96, pp. 2995-3005.
Epub 2006 Aug 30., Dec 2006.
V. K. Mushahwar and K. W. Horch, "Selective activation of muscle
groups in the feline hindlimb through electrical microstimulation of
the ventral lumbo-sacral spinal cord," IEEE Transactions on
Rehabilitation Engineering, vol. 8, pp. 11-21, 2000.
V. K. Mushahwar, D. F. Collins, and A. Prochazka, "Spinal cord
microstimulation generates functional limb movements in chronically
implanted cats," Experimental Neurology, vol. 163, pp. 422-429,
2000.
R. Herman, J. He, S. D'Luzansky, W. Willis, and S. Dilli, "Spinal cord
stimulation facilitates functional walking in a chronic, incomplete
spinal cord injured," Spinal Cord., vol. 40, pp. 65-8, 2002.
M. R. Carhart, J. He, R. Herman, S. D'Luzansky, and W. T. Willis,
"Epidural spinal-cord stimulation facilitates recovery of functional
walking following incomplete spinal-cord injury," IEEE Trans
Neural Syst Rehabil Eng, vol. 12, pp. 32-42, Mar 2004.
C. A. Angeli, V. R. Edgerton, Y. P. Gerasimenko, and S. J. Harkema,
"Altering spinal cord excitability enables voluntary movements after
chronic complete paralysis in humans," Brain, Apr 8 2014.
S. E. Mondello, M. R. Kasten, P. J. Horner, and C. T. Moritz,
"Therapeutic intraspinal stimulation to generate activity and promote
long-term recovery," Front Neurosci, vol. 8, p. 21, 2014.
A. Prochazka, "Chronic techniques for studying neurophysiology of
movement in cats," in Methods for Neuronal Recording in Conscious
Animals (IBRO Handbook Series: Methods in the Neurosciences, Vol.
4), R. Lemon, Ed., ed New York: Wiley, 1984, pp. 113-128.
W. F. Agnew, D. B. McCreery, T. G. H. Yuen, and L. A. Bullara,
"Effects of prolonged electrical stimulation of the central nervous
system," in Neural Prostheses. Fundamental Studies, W. F. Agnew
and D. B. McCreery, Eds., ed Englewood Cliffs, NJ: Prentice Hall,
1990, pp. 225-252.
J. A. Bamford, K. G. Todd, and V. K. Mushahwar, "The effects of
intraspinal microstimulation on spinal cord tissue in the rat,"
Biomaterials.,
vol.
31,
pp.
5552-63.
doi:
10.1016/j.biomaterials.2010.03.051. Epub 2010 Apr 28., Jul 2010.
S. Snow, S. C. Jacobsen, D. L. Wells, and K. W. Horch,
"Microfabricated cylindrical multielectrodes for neural stimulation,"
IEEE Transactions on Biomedical Engineering, vol. 53, pp. 320-6,
2006.
C. Cheng, J. Kmech, V. K. Mushahwar, and A. L. Elias,
"Development of surrogate spinal cords for the evaluation of
electrode arrays used in intraspinal implants," IEEE Trans Biomed
Eng, vol. 60, pp. 1667-76, Jun 2013.
V. R. Edgerton, N. J. Tillakaratne, A. J. Bigbee, R. D. de Leon, and
R. R. Roy, "Plasticity of the spinal neural circuitry after injury," Annu
Rev Neurosci, vol. 27, pp. 145-67., 2004.
S. B. Jun, M. R. Hynd, K. L. Smith, J. K. Song, J. N. Turner, W.
Shain, et al., "Electrical stimulation-induced cell clustering in
cultured neural networks," Med Biol Eng Comput., vol. 45, pp. 101521. Epub 2007 Aug 8., Nov 2007.
S. H. Jeong, S. B. Jun, J. K. Song, and S. J. Kim, "Activity-dependent
neuronal cell migration induced by electrical stimulation," Med Biol
10
[34]
[35]
[36]
[37]
[38]
Eng Comput., vol. 47, pp. 93-9. doi: 10.1007/s11517-008-0426-8.
Epub 2008 Nov 26., Jan 2009.
W. M. Grill, S. E. Norman, and R. V. Bellamkonda, "Implanted
neural interfaces: biochallenges and engineered solutions," Annu Rev
Biomed Eng, vol. 11:1-24., pp. 10.1146/annurev-bioeng-061008124927., 2009.
L. Guevremont and V. K. Mushahwar, "Tapping into the spinal cord
for restoring function after spinal cord injury," in Neural
Engineering: Research, Industry and the Clinical Perspective, D. a.
Bronzino, Ed., ed: CRC Press, 2008, pp. 19-1-26.
L. Guevremont, J. A. Norton, and V. K. Mushahwar, "Physiologically
based controller for generating overground locomotion using
functional electrical stimulation," J Neurophysiol, vol. 97, pp. 2499510, Mar 2007.
K. A. Mazurek, B. J. Holinski, D. G. Everaert, R. B. Stein, R. EtienneCummings, and V. K. Mushahwar, "Feed forward and feedback
control for over-ground locomotion in anaesthetized cats," J Neural
Eng., vol. 9, pp. 026003. doi: 10.1088/1741-2560/9/2/026003. Epub
2012 Feb 13., Apr 2012.
K. A. Mazurek, B. J. Holinski, D. G. Everaert, V. K. Mushahwar, and
R. Etienne-Cummings, "A Mixed-Signal VLSI System for Producing
Temporally Adapting Intraspinal Microstimulation Patterns for
Locomotion," IEEE Trans Biomed Circuits Syst, vol. 9, p. 9, Mar 9
2016.
Jeremy Bamford received the Ph.D.
degree in Neuroscience from the
University of Alberta, Edmonton, AB,
Canada in 2009. Subsequently he
received postdoctoral training in the
Department of Biomedical Engineering
at Duke University, Raleigh, NC, USA.
He is currently an assistant professor in
the Department of Neurosurgery at the
Tulane University School of Medicine,
New Orleans, LA, USA. His current research interests include
intraoperative neurophysiology, especially the mapping of the
vagus nerve during peri-laryngeal surgeries, and the
identification of the central sulcus during tumor resections.
R. Marc Lebel received the B.Sc.
degree in physics from the University of
Calgary, Canada, in 2003, the M.Sc.
degree in medical biophysics from
Western University, Canada, in 2005,
and PhD degree in biomedical
engineering from the University of
Alberta, Canada, in 2010.
He is currently a scientist with GE
Healthcare and adjunct assistant
professor in radiology at the University of Calgary. He
specializes in new MRI acquisition and reconstruction methods.
Kian Parseyan received his bachelor of
science in neuroscience in 2012 from the
University of Alberta in Edmonton,
Canada. He went on to work at Dr.
Mushahwar’s laboratory as a research
assistant with a focus on refining the
fabrication and insertion methods of the
microwires described in this paper. Kian
now leads a private company that is
enabling people to buy and sell homecooked meals.
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TNSRE.2016.2555959, IEEE
Transactions on Neural Systems and Rehabilitation Engineering
TSRNE-2015-00249.R1
Vivian K. Mushahwar (M’97) received
the B.S. degree in electrical engineering
from Brigham Young University, Provo,
UT, in 1991, and the Ph.D. degree in
bioengineering from the University of
Utah, Salt Lake City, in 1996. She
received postdoctoral training at Emory
University, Atlanta, GA, and the
University of Alberta, Edmonton, AB,
Canada. She is currently a Professor in
the Department of Medicine and Centre
for Neuroscience, University of Alberta. Her research interests
11
include identification of spinal cord systems involved in
locomotion,
development
of
spinal-cord-based
neuroprostheses, incorporation of motor control concepts in
functional electrical stimulation applications, and development
of systems for alleviating secondary side effects of immobility
such as pressure ulcers.
Dr. Mushahwar is a member of the IFESS, American
Physiological Society, and Society for Neuroscience. She is the
leader of the Project SMART (Sensory Motor Adaptive
Rehabilitation Technology) team and the Director of the
Canada Foundation for Innovation Centre for Neural Interfaces
and Rehabilitation Neuroscience.
1534-4320 (c) 2016 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.