(K-431) Sputtered Ruthenium Oxide Film for Sub-chronic Neural Recording in Rat Motor Cortex

Introduction:: Intracortical microelectrode arrays (MEAs) have been used to record neural activity in basic science research and for the development of neuroprosthetics. However, their ability to record action potentials is often compromised by the neuroinflammatory response, characterized by blood-brain barrier disruption, glial scar formation, and neuronal loss. Previous work has demonstrated that implantable devices with smaller cross-sectional areas can reduce the formation of the glial scar, improving their ability to record over long periods of time. In addition, the integration of low impedance film materials becomes critical to enhance MEAs' recording capabilities. Current advancements in MEA microfabrication technology involve sputtered iridium oxide (SIROF) to augment the electrochemical surface area while preserving a small GSA, leading to electrodes with reduced electrochemical impedance. Here, we microfabricated ultra-thin devices using sputtered ruthenium oxide (RuOx) as a low impedance, cost-effective film material and investigated its ability to record single neuron activity over a sub-chronic implantation of 6 weeks.

Materials and Methods:: We designed MEAs using amorphous silicon carbide (a-SiC) with 16 electrode sites of distributed across 4 shanks, each with a geometric surface area of 200 µm2. Electrode sites were coated with sputtered RuOx. Six Sprague Dawley rats were subjected to surgical MEA implantation targeting the motor cortex. Then, we recorded neural activity under anesthesia on a weekly basis over a 6-week period post-implantation. Data were filtered using a bandpass between 300 – 3000 Hz and common median referencing to reduce noise and artifacts. Spikes were detected using a thresholding technique of -4σ from the mean and single units were identified using k-means clustering followed by manual validation. The active electrode yield (AEY) was calculated as the percentage of electrode sites detecting single units. Then, we calculated the voltage peak-to-peak (Vpp) of each single unit and the root-mean-square (RMS) noise without detected spikes. The signal-to-noise ratio was calculated by dividing the Vpp by the RMS noise. Finally, we used linear regression to understand the recording changes over the post-implantation time-period. All results are given as mean ± SD.

Results, Conclusions, and Discussions::

We found that right after implantation the AEY was 74.2% and 74.0% six weeks after; however, we observed fluctuations with a lowest AEY of 60.9% at two and four weeks after implantation. The Vpp remained mostly unchanged throughout the six weeks beginning at 128.9±11.3 µV and linear regression showed a loss of less than 1 µV per week. Similarly, the noise levels remained mostly constant, beginning at 9.25±2.99 µV and a slope of -0.13 µV per week. Because of the stability of these two metrics, the SNR remained stable throughout the implantation period beginning at 13.8±6.4 at the time of implantation and showed a decline of 0.18 per week.Throughout the 6-week post-implantation period, sputtered RuOx demonstrated its ability to record neural signals with AEY comparable to devices using SIROF, which have shown AEY between 70-90% in the same timeframe. The quality of the recordings from RuOx was found to be stable with only insignificant losses of signal amplitude and signal-to-noise ratio. These results highlight the reliability of RuOx as a recording material and potential alternative to SIROF, which has recently experienced cost increases due to limited supply. Future studies will investigate the long-term reliability of this material and characterize the neuroinflammatory response to validate its use in neural interfaces.

Acknowledgements (Optional): :

References (Optional): :