A 1024-Channel 10-Bit 36- μW/ch CMOS ROIC for Multiplexed GFET-Only Sensor Arrays in Brain Mapping

Area of Science:

• Neuroscience
• Electrical Engineering
• Materials Science

Background:

• Electrocorticography (ECoG) is crucial for brain mapping.
• Existing neural read-out integrated circuits (ROICs) face scalability and cost challenges.
• Solution-gated GFET sensing probes offer potential for high-density neural interfaces.

Purpose of the Study:

• To develop a scalable, low-power 1024-channel ROIC for GFET-based ECoG.
• To enable cost-effective hybrid headstages for massive neural recordings.
• To facilitate the capture of infra-slow neural signals.

Main Methods:

• Design and fabrication of a 1024-channel ROIC using 0.18 μm CMOS technology.
• Implementation of time-domain multiplexing for GFET-only arrays.
• Integration of low-power CMOS analog frontend with CDS and 10-bit A/D conversion.
• Development of an automated in-situ GFET sensor calibration methodology.

Main Results:

• Fabrication of a compact (0.012 mm²) and low-power (36 μW/channel) 1024-channel ROIC.
• Successful experimental testing with GFET probes on custom FPGA-based headstages.
• Demonstration of infra-slow neural signal recording capabilities.
• Achieved largest scalability in hybrid platforms compared to state-of-art neural ROICs.

Conclusions:

• The developed ROIC enables highly scalable and cost-effective hybrid ECoG platforms.
• The system facilitates the recording of infra-slow neural signals for advanced brain mapping.
• This work advances neural interface technology for large-scale brain activity monitoring.