A very large-scale microelectrode array for cellular-resolution electrophysiology.

Area of Science:

• Neuroscience
• Biomedical Engineering
• Electrical Engineering

Background:

• Traditional electrophysiology faces limitations in channel count and non-invasive interfaces due to inefficient electronics and proximity requirements.
• Scaling to thousands of low-noise recording channels non-invasively is a significant challenge in current electrophysiological studies.

Purpose of the Study:

• To develop a high-density electrode platform enabling simultaneous recording and stimulation of a large number of channels.
• To overcome the limitations of traditional electrophysiology for non-invasive, high-channel-count neural interfaces.

Main Methods:

• Utilized compressed sensing concepts and silicon complementary metal-oxide-semiconductors (CMOS) technology.
• Developed a platform with 65,536 simultaneously recording and stimulating electrodes, each with minimal electronic footprint (25.5 μm x 25.5 μm).
• Implemented a high-performance processing pipeline with a 1 GB/s data rate for real-time analysis.

Main Results:

• Achieved concurrent recording from 65,536 electrodes with approximately 10 µV r.m.s. noise.
• Sensed neural spikes from over 34,000 electrodes across the entire mouse retina.
• Successfully sorted and classified more than 1700 neurons post-visual stimulation.
• Demonstrated stimulation of individual neurons using the electrode array while monitoring network activity.

Conclusions:

• The developed CMOS-based platform significantly advances high-density electrophysiology, enabling unprecedented recording and stimulation capabilities.
• This technology overcomes previous scalability and invasiveness barriers, paving the way for more comprehensive neural interface studies.
• The platform's design and processing pipeline are adaptable to various electrophysiological applications and electrode configurations.