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.