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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

Patient Specific Cortical Electrodes for Sulcal and Gyral Implantation Shayne Morris, Masayuki Hirata, Hisato Sugata, Tetsu Goto, Kojiro Matsushita, Takufumi Yanagisawa, Youichi Saitoh, Haruhiko Kishima, and Toshiki Yoshimine

Abstract—[Purpose] Non-invasive localization of certain brain functions may be mapped on a millimetre level. However, the inter-electrode spacing of common clinical brain surface electrodes still remains around 10 mm. Here, we present details on development of electrodes for attaining higher quality electrocorticographic signals for use in functional brain mapping and brain-machine interface (BMI) technologies. [Methods] We used platinum-plate-electrodes of 1 mm-diameter to produced sheet electrodes after the creation of individualized molds using a 3D-printer and a press system that sandwiched the electrodes between personalized silicone sheets. [Results] We created arrays to fit the surface curvature of the brain and inside the central sulcus, with inter-electrode distances of 2.5 mm (a density of 16 times previous standard types). Rat experiments undertaken indicated no long term toxicity. We were also able to custom design, rapidly manufacture, safely implant and confirm the efficacy of personalized electrodes, including the capability to attain meaningful high gamma-band information in an amyotrophic lateral sclerosis patient. [Conclusion] We developed cortical sheet electrodes with a high spatial resolution, tailormade to match an individual’s brain. [Significance] This sheet electrode may contribute to the higher performance of BMI’s. Index Terms- cortical electrodes, brain machine interface, personalized, brain sulcus, amyotrophic lateral sclerosis (ALS)

I. INTRODUCTION

R

ECENT developments in the area of brain machine interfaces (BMI) have led to the possibility of a seamless interface between the human brain and devices [1]–[7]. To achieve this, neural signals must be attained from the brain and then decoded to interpret their meaning, where decoding relies on the quality of the measured neural signals, particularly those of the high gamma-band range (60 – 200 Hz). One practical method for obtaining high quality neural activity is intracranial electrodes. Several needle electrode arrays have been developed and used on animals and even in human patients [8]–[11], but these electrodes have been reported to be susceptibility to electromagnetic noise, micromotion [12] and to undergo adverse tissue reactions [13]. In addition, needle type electrodes are invasive since they must be inserted into brain tissue. On the other hand, brain surface electrodes or electrocorticographic (ECoG) electrodes are less invasive because they do not penetrate the brain tissue. In comparison with standard electroencephalographic electrodes, ECoG electrodes provide not only superior spatial resolution, but also deliver signals with higher sensitivities in the higher frequency

ranges, making them useful for neural decoding utilizing high gamma band event-related oscillatory changes [14], [15]. We have also shown that intra-sulcal ECoG electrodes on the motor bank side of the central sulcus provided a higher performance than those of the gyral ECoG [14]. However, the inter-electrode spacing of standard clinical type ECoG electrodes used still remains at about 1cm. The surface of the human brain is curved with bumps, depressions and grooves, called gyri and sulci. These structures allow for increased cortical surface area and greater neural processing, and in the case of the central sulcus, with the anterior surface controlling motor function and the posterior surface being responsible for somatosensory function. By inserting electrodes into this sulcus, we may be able to attain more information on movements. Indeed, it has previously been shown that the placement of electrodes into the central sulcus is superior to placement on the brain surface alone from the perspective of decoding accuracy [14]. However, the shape, and depth of the central sulcus is highly individualized, thus making a one-fits-all approach to the design of electrodes, which has been the mainstay until now, inefficient. More recently, high-density electrode arrays that are flexible and foldable in the form of sheets have also been developed [11], [16]–[18]. However, as the majority of these are in the form of sheets, despite being flexible, they may not necessarily provide maximum contact over the irregularly curved surfaces seen within the central sulcus and on the cortical surface. With this in mind, in order to attain neural measurements that are stable over time, that have a higher signal resolution than presently available ECoG electrodes, and that are less invasive than needle type electrodes, we aimed to create a high density array of electrodes designed to match the contour of an individual’s brain surface as well as to place electrodes both on the gyral surface, and inside the central sulcus. In this article, we also briefly describe the placement of patient-specific cortical electrodes into an ALS patient. II. METHODS A. Electrode Sheets 1) Sulcal Electrode Sheet: For the previously stated reasons, we decided to create platinum plate sheet electrodes with an electrode diameter of 1 mm, 1/3rd the diameter of previous standard types, while individualizing the shape of the sheet containing the electrodes using a press mold system. Our method sandwiches the electrodes in-between silicone sheets

Copyright © 2013 IEEE. Personal use of this material is permitted. However, permission to use this material for any other purposes must be obtained from the IEEE by sending an email to [email protected]. 0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014 individualized to fit the surface of an individual patient’s brain. Informed consent was obtained from healthy volunteers and patients, after authorization from the medical ethics committee of our institution. First we took a 1 mm thin-slice MRI series of the subject’s brain. In the case of the model demonstrated in this paper, the first subject was an intractable epileptic patient. Next, we converted the patient’s DICOM data into the NIFI format. We then imported this data into BrainVISA (BrainVISA 4.0.2, http://brainvisa.info/), where we ran the sulci extraction routine which allowed us to generate a 3D image of the brain surface, and also exudate the majority of the sulci (Fig. 1 (a)). Having generated the shape of the various sulci, we selected those making up the central sulcus. After selecting the related pieces, they were exported in the MESH format and were then converted into the STL format. The pieces of the central sulcus in the STL format were then imported into two 3D computer aided designing (CAD) software (Mimics v14.12, 3-matic v5.1, Materialise N.V. Leuven Belgium). Using these software, we combined all the pieces of the central sulcus to a 3-dimentional representation of the central sulcus (Fig. 1 (a)). The created object consists of 2 surfaces, namely, representations of the motor side of the central sulcus and the sensory side. These 2 surfaces were then separated, and the

2 motor side was selected for further processing with 3-matic, because, as was previously stated, the motor side provides superior decoding accuracy of motor function [14], [15]. Next, sides were extended downward from the surfaces, after which, a base was then added to generate CAD data for a mold of the central sulcus. Then an inverse of this mold was produced, so that we now had both male and female molds. Holes were placed in the four corners of the base of the molds to allow for adjustments of the thickness of the silicon electrode sheets. This data was then exported in the STL format (Fig. 1 (a)). Using this data, we then used a 3D printer (Polyjet 3D printer, Objet Geometries Ltd., Israel) to create the actual molds. The contour mold portion representing the central sulcus was 95 mm at maximum length and measured between 6 mm and 24 mm in depth depending on the location along the maximum length axis (Fig. 1 (b)). We marked the desired positions of the electrodes on the mold surface, with the surface facing the motor bank of the central sulcus having 35 electrodes while the surface facing the sensory bank had 15, since the primary motor cortex provides greater information on motor function. We also placed a higher emphasis on the hand-knob area in the form of higher electrode densities.

Fig. 1. (a) Various sulci shown on a 3-dimensional image of the brain, generated with BrainVISA from a subjects MRI data. Sulcus data is removed and converted into the STL format for further processing. CAD data of a mold created from the surface facing the motor side of the central sulcus was created with 3-matic. (b) Silicone sheet compression between male and female molds created with a 3D printer. (c) A silicone sheet is created on the mold, and holes were made for electrode embedding. Flexible stainless leads were then individually connected. A second silicone sheet is adhered to the back of the sheet implanted with the electrodes. Two layers of electrodes are embedded into the silicone sheets, with one layer facing the motor bank, and one layer facing the sensory bank, and a silicone sheet in between to isolate the two electrode arrays.

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

A silicone sheet (SILASTIC MDX4-4210) was then heated to 76°C for 5 seconds and placed on the mold. A heat gun was used to improve contact with the mold surface and to remove air bubbles. The sheet was then allowed to cool naturally to room temperature and left for 3 days to harden. The silicone sheet was then carefully removed from the molds. The same process was repeated to produce 3 sheets (one for the motor side (M1 sheet), one for sensory side (S1 sheet), and one for placement in between these 2 sheets (middle sheet) to isolate the 2 sets of plate electrodes). Next, holes were punched out for the electrodes at predetermined locations in the sheet facing M1. Platinum electrodes (diameter 1 mm, thickness 0.02 mm) were in-set into these locations and temporarily fixed before flexible stainless leads (diameter 0.05 mm polyurethane coated, impedance ≤ 80Ω) were attached individually to each of the electrodes (impedance: electrodes ≤ 10Ω, lead ends (platinum) ≤ 10Ω). After completion of attachments, the lead wires were bundled together and passed through a silicone tube. To fix the electrodes, wires, and silicone tube, a silicone adhesive was added and the second silicone sheet placed on top and compressed in the mold (Fig. 1 (b)). The process was repeated for the sensory side electrodes, the end result being 2 sets of plate electrodes sandwiched between 3 silicone sheets, with the leads all protruding from a single location near the middle of the sheet electrode at the surface of the brain (Fig. 1 (c)). 2) Gyral Electrode Sheet: Using the same process, we were able to not only create sulcal electrodes, but also patientspecific gyral electrodes. These sheets were designed to have electrodes on only the side facing the cortical surface, and the locations and densities of the electrodes were tailored to match the functional areas of this surface. Next, the same 3D printing methods were used to create a model of the subject’s brain to confirm that the manufactured electrodes fitted properly into the central sulcus, and onto the surface of the brain (Fig. 1 (d)). B. Strength, Cytotoxicity and Biocompatibility Tests In order to test the safety of the electrodes before implantation into patients, in accordance with Japanese regulations and requirements for clinical applications, we undertook the following tests. 1) Physical Tests: Prior to clinical trials, to test the strength of the sheet electrodes, a 3 dimensional sheet electrode (10 mm x 60 mm silicone electrode sheet with 50 mm x 1.3 mm diameter lead) was prepared and underwent dynamic tensile and compression tests (TENSILON RTF1250 (A&D), Gauge length 25 mm, stretching speed 500 mm/min, load cell 1kN, 10kN, 22°C 57% humidity). 2) Cytotoxicity Tests: In accordance with the tests required by Japanese regulations for clinical trials, cytotoxicity tests were undertaken with concern to the effects of 3D-high density electrode extract on Chinese hamster fibroblasts (JCRB0603:V79) in accordance with the ISO 10993-5: 2009(E) Biological evaluation of medical devices - Part 5: Tests for in vitro cytotoxicity, ISO 10993-12: 2007 (E) Biological evaluation of medical devices - Part 12 : Sample preparation

3

and reference materials standards, and other related Japanese protocols. In these tests, the 3D grid electrodes were broken down into 2 x 15 mm sections containing all the constituents of the electrodes. Ten mL of medium per 1g of morcellation was added to these sections and then, sealed, shielded from light, and cultivated with the Chinese hamster fibroblasts (JCRB0603:V79) for 24 hours. As positive controls, polyurethane films containing 0.1% and 0.25% zinc diethyldithiocarbamate (referred to respectively, as positive control A and B respectively) were used, while a high density polyurethane film was used as a negative control. 3) Biocompatibility Tests: In accordance with tests required by Japanese regulations for clinical trials, 3D-high density electrodes and control materials (diameter 20 mm) were implanted sub-cutaneously into 12 male rats and then removed after 26 weeks to examine the effects of the subject material on the implanted rats. This study was carried out in accordance with all required and related regulations and protocols for the protection of animals. The electrodes and a control material were symmetrically implanted subcutaneously at the dorsal end of the spinal column while the subject was under intraperitonial anesthesia, with the electrodes implanted on the left side and the control (high density polyethylene sheet (1 mm x 50 mm x 100 mm ) implanted on the right side of the scapula. C. ALS Trial Using the above processes, we also designed 3D personalized electrodes for temporary implantation into an Amyotrophic Lateral Sclerosis (ALS) patient, as part of our clinical brain machine interface trial on severe ALS patients. The patient was a 61 year old male, whom was only able to communicate using eye movement to indicate characters on a letter chart, or by using a personal computer with a system that allowed for movement and clicking of a cursor through a 1 channel input system triggered by minute movements of the patient’s lips. He had total upper and lower limb paralysis and was in a locked-in state. Informed consent was obtained on repeated occasions, and the patient was well aware that the electrode implant would be removed after 21 days, and would have no direct permanent benefit for him himself. The trial was thoroughly reviewed and authorized by the medical ethics committee of our institution. Pre-operative magnetoencephalography (MEG) tests were undertaken to test the suitability of the patient, prior to electrode implantation. In order to avoid possible errors in the determination of the central sulcus, in addition to anatomical determinations, we perform motor tasks studies on functional MRI and somatosensory neuromagnetic evoked field using MEG to confirm its location. III. RESULTS A. Electrode Sheets Using the semi-automated processes mentioned in the methods section, we were able to manufacture 2 types of

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014 patient-specific electrode sheets. These were the central sulcus electrode sheet, and gyral surface electrode sheet. 1) Central Sulcus Electrode Sheet: We were able to manufacture an array of electrodes with an inter-electrode distance between electrodes of 2.5 mm (a density of 16 times that of previous standard types), and we were able to create sheet electrodes molded to fit both the surface curvature of the brain and to fit inside the central sulcus with a total of 50 plate electrodes (35 facing the motor bank and 15 facing the sensory bank). The length of the central sulcus electrode shown is around 95 mm with a maximum depth of close to 24 mm. Unlike the standard-type electrodes that we have previously placed within the central sulcus at our institution which were strip electrodes (4 x 1 electrodes) in one dimension, our patientspecific sheet electrode was 2 dimensional (11~17 x 2~3 electrodes). Distances from the surface of the sheet electrode to the mold were 2.2 mm and 0.1 mm at maximum length extremities, and 3.8 mm and 1.2 mm at maximum width extremities, with an average distance of 1.8±1.4 mm (mean±SD). The thickness varied between 0.81 ~ 0.96 mm at sampled locations. Using a 3D model of the same subject’s brain, also created with the 3D printer, we confirmed that the electrodes fitted into the central sulcus (Fig. 1 (d)). 2) Gyral Surface Electrode Sheet: We also used a similar process to create patient-specific electrodes that matched the contour of the brain's gyral surface. Unlike the central sulcus electrodes which had 2 layers of electrodes, these electrodes are single layered. They may also be placed over the top of the sulci electrodes to cover a predetermined surface of the brain (Fig. 1 (d)). The sheet shown here has 70 electrodes, which are located

4 at various densities in accordance with anatomically perceived functionality, such as the hand-knob. As well as matching the contour of the patient’s cortical surface, the sheet electrodes are softer than standard electrodes and are thus less invasive. B. Strength, Cytotoxicity and Biocompatibility Tests 1) Physical Tests: The electrodes passed the set strength level (silicone sheet 5N, leads 50N) tests with values of 9.58N for the silicone sheet, and more than 60N for the leads. 2) Cytotoxicity: A comparison with the extract containing the electrode extract, and the controls was made. Using the 100% negative control extract we confirmed that the number of colonies were almost unchanged. The IC50s for Positive Control A and B were 0.91% and 57.0% respectively. The IC50 for the positive control was 2.31μg/mL, thus indicating no strong cytotoxicity on colony growth. (Supplementary Figure 1). 3) Biocompatibility: One out of 12 rats showed slight swelling at the sites of implantation four days after subcutaneous placement. In this subject, swelling on the electrode implanted side continued for 41 days, while swelling on the control implanted side was also observed for 19 days. Superficial abnormalities disappeared after 20 days on the control implanted side and 42 days on the electrode implanted side but reoccurred on the electrode implanted side 175 days after implantation. Examinations after the completion of the test period indicated the presence of an abscess most likely due to infection at the time of implantation. None of the other 11 subjects developed abnormalities over the test period.

Fig. 2. (a) CAD model of ALS patient’s brain surface with pink area indicating minimal target area covering pre- and post- central gyri. (b) Completed sheet electrodes placed on a brain surface model. CS indicates the central sulcus, and HK indicates the hand knob. (c) Pre-operative imaging for intra-operative navigation indicating locations of pre-central gyrus (magenta) and major cortical veins (red) etc. Blue dotted line indicates planned craniotomy. (d) Intra-operative image after placement of electrodes, with the yellow dotted line indicating the central sulcus (CS) and the hand knob (HK) area indicated in blue.

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

5

Fig. 3. (a) Intra-operative EEG monitoring confirming signals attained from all electrodes. (b) Data during imagined hand movement. Increased amplitude in the high gamma band prior to movement onset is observed particularly over the hand knob (HK) area in a paralyzed ALS patient. Dotted black line indicates the central sulcus (CS).

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

6

C. ALS Trial In this patient, we designed only gyral electrodes. After the creation of gyral molds using our patient-specific method described previously (Fig. 2 (a)), sheet electrodes were manufactured targeting mainly the primary motor cortex, primary somatosensory cortex, and the premotor cortex (53 plate electrodes over the primary motor cortex (with a higher density over the hand knob), 17 over the premotor cortex, and 24 over the somatosensory cortex, with 1 subdural reference plate electrode facing toward the dura mater, and a ground electrode for epidural placement). Leads from the plate electrodes were gathered into 2 groups, coated in a silicon sheath, and extended out from 2 locations on the sheet (Fig. 2 (b)). It took 1 month to complete the manufacturing of the electrodes (2 days to finalize the CAD design, 14 days to manufacture the mold, including shipping, and 14 days to manufacture the final sheet electrode). We used a neuronavigational system for pre-surgical planning and intraoperative navigation. We pre-operatively located the precentral gyrus and surrounding major cortical veins (Fig. 2 (c)), and determined the optimal site for craniotomy. Intraoperatively, we performed craniotomy at the predetermined site, and identified the pre-operative anatomical features of the cortical surface (pre-central gyrus, and cortical veins etc.) after the dural membrane was opened. We also reconfirmed the location the central sulcus through N20 phase reversal of somatosensory evoked potentials. The sheet electrode was placed on arachnoid surface of the left hemisphere in the patient based on these pre-operative and intra-operative findings. The electrodes closely fitted the contour of the brain surface as was designed (Fig. 2 (d)). Intra-operatively monitoring showed clear ECoG signals from all electrodes (Fig. 3 (a)). In post-operative ECoG tests, motor imaginary tasks induced localized high gamma band activity in the hand-knob area (Fig. 3 (b)). The electrodes were removed as scheduled 21 days after implantation without any further deterioration in neurological symptoms. IV. DISCUSSION In the present study, by using semi-automated methods to map sulci and by utilizing CAD software and 3D printers, we created patient-specific cortical sheet electrodes with high spatial resolution for not only gyral but also sulcal implantation. Using our method, we are able to rapidly design, manufacture, and implant electrodes, as well as obtain signals from every electrode, including high frequency information after temporary implantation into an ALS patient. In order to reduce production times, we used a semiautomated process to design electrodes for placement into the central sulcus and on to the gyral surface. We used a 3D printer for the rapid manufacturing of molds that were in turn used to produce patient-specific high-density sheet electrodes capable of covering a wide anatomical area with ability to set higher electrode densities over functionally significant areas such as the hand knob. Through the use of this approach, we were able to manufacture patient-specific high density cortical electrodes

within 1 month for use in a clinical BMI trial on an ALS patient. Using our method, we were able to manufacture a sheet electrode array for placement into the central sulcus. Due to its patient-specific design, the electrode is less invasive than the standard sheet electrodes used in the past, and provides a higher dimensional array of electrodes than the strip electrodes previously utilized. Electrode locations can also be altered to provide higher densities within the central sulcus over areas such as the hand knob, and can be placed over the whole length of the central sulcus. Our electrodes also allow for electrodes to be placed in the direction of the motor bank, as well as the somatosensory bank, which may provide the opportunity for the sensory feedback required in close-loop systems [19], [20]. Our central sulcus electrode was designed to be simultaneously used in conjunction with our gyral electrodes. Using the above method, we were also able to manufacture a sheet electrode for placement over the gyral surface. These sheet electrodes are patient specific and designed to match the individual surfaces of a patient’s brain, thus meaning they apply less pressure at localized areas of contact than the previous types of flat 2-dimensional sheet electrodes, because all of the electrodes are in contact with the brain surface. This also means that compared to flat electrode sheets, with larger type plate electrodes, there is a higher probability that the plate electrodes on our sheet electrodes will be in contact with the brain surface, thus there is a greater chance of attaining useful EEG from all electrodes (Supplementary Figure 2). Due to these characteristics, the electrodes can be designed to cover relatively large areas of the cortical surface, and thus have the potential to simultaneously record data on a wide range of brain functions. It is also possible to increase the density of electrodes over cortical areas of higher importance, for example the hand knob. In our ALS patient trial, readable signals were obtained from all plate electrodes on the implanted sheet. It is noteworthy that we were successfully able to gain gamma-band information from specific functional areas such as the hand knob. Our high density electrodes successfully detected a variety of gamma band activities that differed from electrode to electrode despite electrodes being in close proximity (approximately 3 mm interelectrode spacing) to one another, indicating that a higher density of electrodes may contribute to higher performance BMI systems. The power of high frequency gamma-band information is weak and it is difficult measure this range using standard forms of scalp EEG or MEG. On the other hand we have previously shown that gamma-band activity is useful for predicting motor activity using standard sized cortical electrodes [15], [21]. Using the high-density patient-specific cortical electrodes described in this study, we were able to measure gamma-band activity with high spatial resolution even in an ALS patient with complete upper and lower limb paralysis, which we believe will allow for the more accurate real-time control of external peripherals including robot devices. Regarding safety, we have successfully completed the nonclinical studies necessary to move on to clinical trials, including those related to physical strength, cytotoxicity, and

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

7

biocompatibility, as well as long term stability. In our ALS patient clinical trial we implanted our sheet electrode for 21 days, with no new neurological symptoms incurred after electrode removal. We will continue clinical evaluations to confirm the safety and efficacy of these sheet electrodes, moving towards the implantation of a fully-implantable wireless system [22]. One limitation of personalized electrodes, particularly in the case of intra-sulcal electrodes, is the problem of cortical surface adhesion and the presence of vascular structures that may make placement of predetermined single electrode arrays within the central sulcus difficult. In the past, in our experience with the placement of standard type electrodes in the central sulcus of chronic pain patients, we have found that fine microscopic dissection by a skilled neurosurgeon can overcome this problem in many cases. In addition, we have found that careful examination of MRI can provide a good understanding of possible vascular obstacles, which can then be taken into account at the planning stage. In some cases, the sheet electrode may be divided to 2 or more sheets. Also, it is possible to produce the electrodes in a sterile environment that would allow for a certain degree of fine-tuning during placement, although this should be used as a last resort. In addition, the brains of many of the ALS patients, who may potentially benefit from our fully-implantable wireless system, are generally atrophic, making implantation into the central sulcus easier. With increasingly improved MRI imaging, we are able to attain greater information of localized anatomical structures. In situations where it is determined intra-sulcal electrodes cannot be safely used, however, placement may be limited to gyral electrodes. V. CONCLUSIONS We developed a high-density patient-specific sheet electrode that fit within the sulci or on gyral surfaces. They are designed and rapidly manufactured with semi-automatic methods involving the use of a 3D-printer and resulted in the effective attainment of ECoG signals. We temporarily placed this patient-specific sheet electrode into an ALS patient and were able to attain readable signals from all plate electrodes and measured localised high gamma band activity in the hand knob area. We believe that this sheet electrode will contribute to higher performance BMI systems ACKNOWLEDGMENTS This study is the result of research funded in part by a Brain Machine Interface Development and Development of BMI Technologies for Clinical Application grant carried out under the Strategic Research Program for Brain Sciences (SRPBS) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (MEXT) , by a KAKENHI (23390347, 26282165) grant funded by the Japan Society for the Promotion of Science (JSPS), and through a Health Labour Sciences Research Grant (23100101) from the Ministry of Health Labour and Welfare. The authors wish to thank Annuar Khairul of Materialize Japan Co. Ltd, YOKOHAMA, JAPAN for his help

and advice with among other things, the manufacturing of the models and casts using the 3D printer. We also wish to thank Shinichi Morikawa and Yoshihiro Watanabe of UNIQUE MEDICAL CO.LTD., OSAKA, JAPAN for their help in the manufacturing process of the sheet electrodes. REFERENCES [1] J. N. Mak and J. R. Wolpaw, “Clinical Applications of Brain-Computer Interfaces: Current State and Future Prospects.,” IEEE Rev. Biomed. Eng., vol. 2, no. 1, pp. 187– 199, 2009. [2] D. J. McFarland, D. J. Krusienski, W. A. Sarnacki, and J. R. Wolpaw, “Emulation of computer mouse control with a noninvasive brain-computer interface.,” J. Neural Eng., vol. 5, no. 2, pp. 101–110, 2008. [3] F. Galán, M. Nuttin, E. Lew, P. W. Ferrez, G. Vanacker, J. Philips, H. Van Brussel, and J. R. Millán, “An asynchronous and non-invasive brain-actuated wheelchair,” in 13th International Symposium on Robotics Research Hiroshima Japan, 2007, pp. 1–11. [4] I. Iturrate, J. Antelis, and J. Minguez, “Synchronous EEG Brain-Actuated Wheelchair with Automated Navigation,” Pattern Recognit., pp. 2318–2325, 2009. [5] F. Cincotti, D. Mattia, F. Aloise, S. Bufalari, G. Schalk, G. Oriolo, A. Cherubini, M. G. Marciani, and F. Babiloni, “Noninvasive brain-computer interface system: towards its application as assistive technology.,” Brain Res. Bull., vol. 75, no. 6, pp. 796–803, 2008. [6] M. A. Lebedev and M. A. L. Nicolelis, “Brain-machine interfaces: past, present and future.,” Trends Neurosci., vol. 29, no. 9, pp. 536–546, 2006. [7] G. R. Muller-Putz, R. Scherer, G. Pfurtscheller, and R. Rupp, “EEG-based neuroprosthesis control: a step towards clinical practice,” Neurosci. Lett., vol. 382, no. 1–2, pp. 169– 174, 2005. [8] E. M. Maynard, C. T. Nordhausen, and R. A. Normann, “The Utah intracortical Electrode Array: a recording structure for potential brain-computer interfaces,” Electroencephalogr. Clin. Neurophysiol., vol. 102, no. 3, pp. 228–239, 1997. [9] P. J. Rousche and R. a Normann, “Chronic recording capability of the Utah Intracortical Electrode Array in cat sensory cortex.,” J. Neurosci. Methods, vol. 82, no. 1, pp. 1–15, Jul. 1998. [10] A. S. Dickey, A. Suminski, Y. Amit, and N. G. Hatsopoulos, “Single-unit stability using chronically implanted multielectrode arrays.,” J. Neurophysiol., vol. 102, no. 2, pp. 1331–9, Aug. 2009. [11] L. R. Hochberg, D. Bacher, B. Jarosiewicz, N. Y. Masse, J. D. Simeral, J. Vogel, S. Haddadin, J. Liu, S. S. Cash, P. van der Smagt, and J. P. Donoghue, “Reach and grasp by people with tetraplegia using a neurally controlled robotic arm.,” Nature, vol. 485, no. 7398, pp. 372–5, May 2012. [12] A. Gilletti and J. Muthuswamy, “Brain micromotion around implants in the rodent somatosensory cortex.,” J. Neural Eng., vol. 3, no. 3, pp. 189–95, Sep. 2006. [13] V. S. Polikov, P. a Tresco, and W. M. Reichert, “Response of brain tissue to chronically implanted neural

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

8

electrodes.,” J. Neurosci. Methods, vol. 148, no. 1, pp. 1–18, Oct. 2005. [14] T. Yanagisawa, M. Hirata, Y. Saitoh, A. Kato, D. Shibuya, Y. Kamitani, and T. Yoshimine, “Neural decoding using gyral and intrasulcal electrocorticograms.,” Neuroimage, vol. 45, no. 4, pp. 1099–106, May 2009. [15] T. Yanagisawa, M. Hirata, Y. Saitoh, T. Goto, H. Kishima, R. Fukuma, H. Yokoi, Y. Kamitani, and T. Yoshimine, “Real-time control of a prosthetic hand using human electrocorticography signals.,” J. Neurosurg., vol. 114, no. 6, pp. 1–8, Feb. 2011. [16] T. Stieglitz, “Flexible biomedical microdevices with double-sided electrode arrangements for neural applications,” Sensors Actuators A Phys., vol. 90, no. 3, pp. 203–211, May 2001. [17] B. Rubehn, C. Bosman, R. Oostenveld, P. Fries, and T. Stieglitz, “A MEMS-based flexible multichannel ECoGelectrode array.,” J. Neural Eng., vol. 6, no. 3, p. 036003, Jun. 2009. [18] J. Viventi, D.-H. Kim, L. Vigeland, E. S. Frechette, J. a Blanco, Y.-S. Kim, A. E. Avrin, V. R. Tiruvadi, S.-W. Hwang, A. C. Vanleer, D. F. Wulsin, K. Davis, C. E. Gelber, L. Palmer, J. Van der Spiegel, J. Wu, J. Xiao, Y. Huang, D. Contreras, J. a Rogers, and B. Litt, “Flexible, foldable, actively multiplexed, high-density electrode array for mapping brain activity in vivo.,” Nat. Neurosci., vol. 14, no. 12, pp. 1599–605, Dec. 2011. [19] Z. Li, J. E. O’Doherty, M. A. Lebedev, and M. A. L. Nicolelis, “Adaptive decoding for brain-machine interfaces through Bayesian parameter updates.,” Neural Comput., vol. 23, no. 12, pp. 3162–204, Dec. 2011. [20] A. L. Orsborn and J. M. Carmena, “Creating new functional circuits for action via brain-machine interfaces.” Front. Comput. Neurosci., vol. 7, p. 157, Jan. 2013. [21] T. Yanagisawa, M. Hirata, Y. Saitoh, H. Kishima, K. Matsushita, T. Goto, R. Fukuma, H. Yokoi, Y. Kamitani, and T. Yoshimine, “Electrocorticographic control of a prosthetic arm in paralyzed patients.,” Ann. Neurol., vol. 71, no. 3, pp. 353– 61, Mar. 2012. [22] M. Hirata, K. Matsushita, T. Suzuki, T. Yoshida, F. Sato, S. Morris, T. Yanagisawa, T. Goto, M. Kawato, and T. Yoshimine, “A Fully-Implantable Wireless System for Human Brain-Machine Interfaces Using Brain Surface Electrodes: WHERBS,” IEICE Trans. Commun., vol. E94-B, no. 9, pp. 2448– 2453, 2011. Shayne Morris MBA M.D. received his MBA from Yokohama National University in 1994. Afterwards, he received an M.D. degree in Medicine from Osaka University in 2002, was a Ph.D. student at Osaka University Graduate School of Medicine and is a neurosurgeon at Nishinomiya Prefectural Hospital in Kobe, Japan.

Masayuki Hirata B.S., M.S. and M.D. and Ph.D. graduated from the Faculty of Engineering at the University of Tokyo in 1985 and Osaka University Medical School in 1994. He is a board-certified neuro-surgeon who specializes in functional neurosurgery. He was promoted to Specially Appointed Associate Professor at the Dept. of Neurosurgery, Osaka University Medical School and is serving as the leader of a neural engineering group. Hisato Sugata Ph.D. received his MSc

degree from Kobe University in 2009, and received his Ph.D. degree in Health Science from Osaka University Graduate School of Medicine in 2012. He is currently working on brain-machine interfaces using MEG at the Department of Neurosurgery, Osaka University Medical School. Tetsu Goto, M.D., Ph.D. graduated from Osaka University Medical School in 2002. He is a board-certified neurosurgeon and is currently working as a neurosurgeon at Kawachi Sogo Hospital in Osaka, Japan. He specializes in the analysis of motor and language signals recorded with MEG and ECoG. Kojiro Matsushita, B.S., Ph.D. graduated from the Dept. of Engineering, Tokyo University of Science in 2000. He graduated from the Graduate School of Engineering, The Univ. of Tokyo in 2007. He was a postdoctoral fellow in the Computer Science and Artificial Intelligence Laboratory at MIT in 2008. He was recruited as a Specially Appointed Assistant Professor at the Department of Neurosurgery, Osaka University Medical School in 2009. Takufumi Yanagisawa M.D., Ph.D. graduated from Dept. of Physics, Waseda University and Osaka University Medical School in 2005. He is a board-certified neurosurgeon. He is currently a Specially Appointed Researcher in the Department of Neurosurgery, Osaka University Medical School, with a principal interest in ECoG decoding.

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also 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/TBME.2014.2329812, IEEE Transactions on Biomedical Engineering

TBME-00158-2014

9

Youichi Saitoh, M.D., Ph.D. graduated from Osaka University Medical School in 1982. He is a board-certified neurosurgeon and is currently a Specially Appointed Professor at the Department of Neuromodulation and Neurosurgery, Osaka University. He specializes in functional neurosurgery; in particular, motor cortex stimulation and deep-brain stimulation. Haruhiko Kishima, M.D., Ph.D. graduated from Osaka University Medical School in 1991. He is a boardcertified neurosurgeon and is currently an Associate Professor at the Department of Neurosurgery, Osaka University Medical School. He specializes in functional neurosurgery, specifically in epileptic surgery and deep brain stimulation. Toshiki Yoshimine, M.D., Ph.D. graduated from Osaka University Medical School in 1975. He is the Director of the Japan Neurosurgical Society. He currently serves as the Professor and Chairman of the Dept. of Neurosurgery, Osaka University Medical School and Director of the Medical Center for Translational Research, Osaka University Hospital.

0018-9294 (c) 2015 IEEE. Translations and content mining are permitted for academic research only. Personal use is also permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.

Patient-specific cortical electrodes for sulcal and gyral implantation.

Noninvasive localization of certain brain functions may be mapped on a millimetre level. However, the interelectrode spacing of common clinical brain ...
990KB Sizes 6 Downloads 3 Views