Development of a Microscale Implantable Neural Interface (MINI) Probe System.

Cortical recording devices hold promise for providing augmented control of neuroprostheses and brain-computer interfaces in patients with severe loss of motor function due to injury or disease. This paper reports on the preliminary in vitro and in vivo results of our microscale implantable neural interface (MINI) probe system. The MINI is designed to use proven components and materials with a modular structure to facilitate ongoing improvements as new technologies become available. This device takes advantage of existing, well-characterized Michigan probe technologies and combines them to form a multichannel, multiprobe cortical assembly. To date, rat, rabbit, and non-human primate models have been implanted to test surgical techniques and in vivo functionality of the MINI. Results demonstrate the ability to form a contained hydrostatic environment surrounding the implanted probes for extended periods and the ability of this device to record electrophysiological signals with high SNRs. This is the first step in the realization of a cortically-controlled neuroprosthesis designed for human applications.

A low-profile three-dimensional silicon/parylene stimulating electrode array for neural prosthesis applications.

This paper describes a low-profile three-dimensional silicon/parylene microelectrode array as basis for practical neural prostheses for use in the central nervous system. The circuit areas of the silicon probes, containing mixed-signal CMOS circuitry for neural stimulation/recording, can be folded over to reduce the overall height of the microassembled array above the cortical surface. The low- profile structure is implemented using multiple gold beams spaced by orthogonal silicon braces. An integrated silicon/parylene batch process is introduced to encapsulate these interconnects and achieve high yield.

Surface modification of neural recording electrodes with conducting polymer/biomolecule blends.

The interface between micromachined neural microelectrodes and neural tissue plays an important role in chronic in vivo recording. Electrochemical polymerization was used to optimize the surface of the metal electrode sites. Electrically conductive polymers (polypyrrole) combined with biomolecules having cell adhesion functionality were deposited with great precision onto microelectrode sites of neural probes. The biomolecules used were a silk-like polymer having fibronectin fragments (SLPF) and nonapeptide CDPGYIGSR. The existence of protein polymers and peptides in the coatings was confirmed by reflective microfocusing Fourier transform infrared spectroscopy (FTIR). The morphology of the coating was rough and fuzzy, providing a high density of bioactive sites for interaction with neural cells. This high interfacial area also helped to lower the impedance of the electrode site and, consequently, to improve the signal transport. Impedance spectroscopy showed a lowered magnitude and phase of impedance around the biologically relevant frequency of 1 kHz. Cyclic voltammetry demonstrated the intrinsic redox reaction of the doped polypyrrole and the increased charge capacity of the coated electrodes. Rat glial cells and human neuroblastoma cells were seeded and cultured on neural probes with coated and uncoated electrodes. Glial cells appeared to attach better to polypyrrole/SLPF-coated electrodes than to uncoated gold electrodes. Neuroblastoma cells grew preferentially on and around the polypyrrole/CDPGYIGSR-coated electrode sites while the polypyrrole/CH(3)COO(-)-coated sites on the same probe did not show a preferential attraction to the cells. These results indicate that we can adjust the chemical composition, morphology, electronic transport, and bioactivity of polymer coatings on electrode surfaces on a multichannel micromachined neural probe by controlling electrochemical deposition conditions.

A multichannel neural probe for selective chemical delivery at the cellular level.

A bulk-micromachined multichannel silicon probe capable of selectively delivering chemicals at the cellular level as well as electrically recording from and stimulating neurons in vivo has been developed. The process buries multiple flow channels in the probe substrate, resulting in a hollow-core device. Microchannel formation requires only one mask in addition to those normally used for probe fabrication and is compatible with on-chip signal-processing circuitry. Flow in these microchannels has been studied theoretically and experimentally. For an effective channel diameter of 10 microns, a channel length of 4 mm, and water as the injected fluid, the flow velocity at 11 torr is about 1.3 mm/s, delivering 100 pl in 1 s. Intermixing of chemicals with the tissue fluid due to natural diffusion through the outlet orifice becomes significant for dwell times in excess of about 30 min, and a shutter is proposed for chronic use. The probe has been used for acute monitoring of the neural responses to various chemical stimuli in guinea pig superior and inferior colliculus.

Silicon ribbon cables for chronically implantable microelectrode arrays.

This paper describes the design, fabrication, and testing of miniature ultraflexible ribbon cables for use with micromachined silicon microprobes capable of chronic recording and/or stimulation in the central nervous system (CNS). These interconnects are of critical importance in reliably linking these microelectrodes to the external world through a percutaneous connector. The silicon cables allow the realization of multilead, multistrand shielded local interconnects that are extremely flexible and yet strong enough to withstand normal handling and surgical manipulation. Cables 5 microns thick, 1-5 cm long, and from 60 to 250 microns wide have been fabricated with up to eight leads. The series lead resistance is typically 4 k omega/cm for polysilicon and 500 omega/cm for tantalum, with shunt capacitance values of 5-10 pF/cm and an interlead capacitance below 10 fF/cm. Soak tests in buffered saline performed under electrical and mechanical stress have been underway for over three years and show subpicoampere leakage levels. Silicon microprobes with built-in ribbon cables have remained functional for up to one year in the guinea pig CNS, recording driven single-unit activity and maintaining impedance levels in the 1-7 M omega range.

Evaluation of a silicon-substrate modiolar eighth nerve implant in a guinea pig.

The availability of thin-film multichannel electrodes provides new possibilities for implantation and direct stimulation of the modiolar portion of the auditory nerve. Electrodes in direct contact with the auditory nerve should be functional at lower thresholds and might require fewer remaining neurons for stimulation compared to electrodes in the scala tympani. This strategy would also provide close contact with neural elements subserving a greater frequency range. Implantation of the eighth nerve may also be advantageous in profoundly deaf subjects who lack an implantable scala tympani. In this study we evaluated the effect of surgical implantation and chronic placement of a silicon substrate implant in the modiolar portion of the auditory nerve of the guinea pig. Of six chronically implanted ears, five showed changes limited to the loss of spiral ganglion cells in the canal of Rosenthal, immediately adjacent to the implant. The sixth ear showed more extensive cochlear alteration in a pattern suggestive of vascular injury. In separate acute experiments, implants were placed in the modiolar portion of the auditory nerve and electrophysiologic analysis was performed. Middle latency responses with good morphology were obtained at thresholds below those found with scala tympani implants. Input-output functions exhibited a plateau in response amplitude at stimulus levels below thresholds for seventh or vestibular portion of the eighth nerve. Further modifications of the modiolar portion of the auditory nerve electrode design will include development of an electrode interconnect that will allow chronic implantation with stimulation.

Strength characterization of silicon microprobes in neurophysiological tissues.

Strength characteristics of thin-silicon probes in neural tissues have been determined experimentally. It is shown that by proper selection of the substrate length, width, and thickness, silicon substrates can be designed and used to penetrate a variety of biological tissues without breakage or excessive dimpling. Thin-silicon structures have a maximum fracture stress which is a factor of six larger than bulk silicon, and are very flexible and capable of bending to angles larger than 90 degrees. Silicon substrates 15 microns thick x 30 microns wide can easily penetrate guinea pig and rat pia arachnoid layers with minimum dimpling and no breakage, while substrates 30 microns thick x 80 microns wide can penetrate guinea pig and rat dura mater repeatedly without breakage. Quantitative comparison on the relative toughness of neurophysiological tissues in rat and guinea pig have also been experimentally obtained.

Batch-fabricated thin-film electrodes for stimulation of the central auditory system.

Silicon micromachining and thin-film technology have been employed to fabricate iridium stimulating arrays which can be used to excite discrete volumes of the central nervous system. Silicon multichannel probes with thicknesses ranging from 1 to 40 microns and arbitrary two-dimensional shapes can be fabricated using a high-yield, circuit-compatible process. Iridium stimulating sites are shown to have similar characteristics to iridium wire electrodes. Accelerated pulse testing with over 8 million 100 microA biphasic current pulses on 8000 microns 2 sites has demonstrated the long-term stability of iridium and activated iridium sites. In vivo tests have been performed in the central auditory pathways to demonstrate neural activation using the devices. These tests show a selective activation both as a function of site separation and site size.