A 4096 channel event-based multielectrode array with asynchronous outputs compatible with neuromorphic processors.

Area of Science:

• Bioelectronics
• Neurotechnology
• Signal Processing

Background:

• Traditional bio-signal sensing methods often require high sampling rates, leading to excessive data and energy consumption.
• Efficient and low-power bio-signal acquisition is crucial for advancements in medical bioelectronics and wearable devices.

Purpose of the Study:

• To develop a novel self-clocked microelectrode array (MEA) for event-based bio-signal digitization.
• To significantly reduce off-chip data transmission and energy consumption in bio-signal sensing.

Main Methods:

• Designed and fabricated a 64x64 microelectrode array with integrated asynchronous pixel-level digitization.
• Implemented an asynchronous 2D-arbiter and Address-Event Representation (AER) communication block for efficient data routing.
• Utilized electrogenic cells for experimental validation and chip characterization.

Main Results:

• The novel MEA digitizes bio-signals at the pixel level, generating asynchronous digital address-events only when signal changes exceed a threshold.
• Demonstrated significant reduction in off-chip data transmission compared to traditional high-sampling-rate methods.
• Successfully interfaced the MEA with a mixed-signal neuromorphic processor for end-to-end event-based sensing and processing.

Conclusions:

• The self-clocked MEA offers a power-efficient and data-efficient approach to bio-signal sensing.
• This event-based sensing paradigm is suitable for real-time monitoring and processing of biological signals.
• The developed system represents a promising prototype for next-generation neuromorphic bioelectronic interfaces.