FABRICATION OF HOLLOW SUBSTRATES AND COMPLIANT RIBBON CABLES FOR NEURAL PROBE WITH INTERNAL ELECTRODES

Neural engineering uses sophisticated probes, consisting of metal electrodes on thin solid substrates, to record neural signals in the brain. They have been most suitable for acute, short-term implants to study the neuronal response to electrical, chemical, and optical stimulation. This approach has been critical to understanding the processes of memory and learning and how they are usurped by addiction and the diseases of the brain. However, these probes begin to fail soon after implantation due to electrode corrosion, cracking of the insulation, and the immune response of the brain which forms a protective sheath on the probe, hence making the recording sites useless. The absence of a probe technology suitable for long term, chronic, implantations is a major limitation for neuro-prosthetics, which must function for the natural life of the patient. The trigger for this immune response, which is still not well understood, has been attributed to breaching of the blood-brain barrier during implantation and abrasion of soft neural tissue by these stiff probes as the brain moves due to breathing, heartbeat and other natural processes. Extensive research has, thus far, failed to solve the chronic implant problem.

This thesis explores the development of several novel hollow substrates for the fabrication of neural probes with integrated thin film conductor wiring. The new probe design is based on the neurotrophic electrode developed by Kennedy et al. [1-4] over the past 25 years. The Kennedy device is a conical glass structure, a few millimeters in length, tapering from 0.5 mm diameter at the base to ~50 µm at the tip. There is a small opening at the tip through which axons, blood vessels and other neurites can infiltrate the cone. The cone is filled with a neurotrophic fluid just prior to implantation which promotes the growth of axons through this opening. The signaling of neurons near the tip is carried along these axons into the cone where it is sensed by fine gold wire electrodes. Brain-computer interfaces based on these probes have functioned reliably for more than 10 years.

Using nanotechnology and lithography techniques, we have been successful in developing various kinds of hollow substrates which can be used to fabricate electrodes on the inside of the structure, which will allow the smaller neurites (required for recording signals) to enter but will block the larger brain cells responsible for reactive immune response. An internal electrode design also isolates electrodes from the chatter of distant neurons, the principal noise source in neural recordings. These advances will enable dramatic improvements in both the spatial resolution and the number of recording channels in these devices.