Technological and Computational Approaches for Large Count High-Density Neural Probes

PASSCODE: MEMS

Implantable neural probes are widely used to study the brain by recording its electrical and chemical responses. The demand to map more neurons has led to the development of new technologies, but there are challenges. Innovative microfabrication techniques are required to create probes with user-defined features like density, size, shape, and distribution. Ensuring that individual probe shanks are minimally invasive and suitable for long-term measurements is another significant challenge. Packaging and integrating these high-count, high-density probes also pose difficulties. Lastly, the computational challenge of high-channel neural recording involves the need for automatic algorithms to cluster large datasets into spike and non-spike classes, as manual sorting becomes impractical.

In this study, we introduce technological microfabrication and packaging techniques to create high-count high-density planar 2D, and non-planar 3D neural probes with user-defined features. We have also optimized the mechanical and electrical features of these probes to enhance functionality while minimizing invasiveness for long-term measurements. The robustness, insertion, and recording functionality of the electrodes have been validated through mechanical and electrical in-vitro tests, demonstrating their suitability for both acute and long-term applications. Additionally, we have developed computational solutions to improve the reliability of spike sorting results using dynamic calculations of the quality metrics.

CHAIR: Professor Khalil Najafi