Development of Chronic Implantable Electrodes for Long-term Visual Evoked Potential Recording in Rabbits

Mariko Kuwabara, Hiroyuki Tashiro, Yasuo Terasawa, Koji Osawa, Takashi Tokuda, Jun Ohta, Takashi Fujikado
Vol. 6 (2017) p. 59-67

Development of direct neural interface (DNI) including visual prostheses absolutely requires confirmation of their long-term safety and stability. Functional evaluation by electrically evoked potentials (EEPs) is effective in this regard, although the recording system must be stable for chronic use. In addition, control of anesthetic depth is important for stable recording of the evoked potentials. The purpose of this study was to develop a chronically implanted electrode capable of recording visual evoked responses safely during repeated anesthesia over long periods, which would allow more effective safety evaluations of not only visual prostheses but also DNI. We developed two types of electrodes, and implanted them into rabbits. A general screw electrode was used for comparison with the novel electrodes. Structurally, the newly developed platinum (Pt) ball-tip screw electrode consisted of a plastic screw with smoothly surfaced Pt balls on the tip. The depth of implantation into the brain was adjustable via a threaded insert installed in the skull. The newly developed platinum/iridium (Pt/Ir) ball-tip planar multi-electrode array (MEA) comprised Pt/Ir ball electrodes placed in a two-dimensional lattice pattern, which was implanted just beneath the skull. These electrodes recorded variations in visual evoked potentials (VEPs) in response to 20 J flash stimuli over a period of 48 weeks. After 48 weeks of implantation, the ability of the electrodes to continue recording EEPs was confirmed (500 µA, 500 µs, cathodic first biphasic). During the recording of VEPs and EEPs, stable anesthesia was maintained with isoflurane (end-tidal 2.4%). The depth of anesthesia using isoflurane could be adjusted safely, and allowed stable recording of evoked potentials throughout the long-term study. However, stable recording using the general screw electrode was possible only for a short period. We also obtained stable latency and N1 amplitude readings over the 48 weeks using the newly developed electrodes, and successfully recorded EEPs after the 48-week period. These results suggest that the novel electrodes work well over the entire duration of the study, and may allow assessment of long-term safety and stability of not only visual prostheses, but also other devices utilizing brain machine interfaces or direct neural interfaces.

Supplementary materials