An energy-efficient micropower neural recording amplifier.

Area of Science:

• Biomedical Engineering
• Neuroscience Instrumentation
• Low-Power Integrated Circuits

Background:

• Neural recording requires amplifiers with high sensitivity and low power consumption.
• Existing amplifiers often struggle to balance noise performance with energy efficiency.
• Common-mode noise rejection is crucial for reliable neural signal acquisition.

Purpose of the Study:

• To develop and characterize an ultralow-power neural recording amplifier.
• To achieve near-theoretical input-referred noise performance.
• To demonstrate the amplifier's capability for recording neural spikes and local field potentials (LFPs).

Main Methods:

• Designed a neural amplifier utilizing low-noise techniques and a differential input stage.
• Configured the amplifier for two distinct operating modes: neural spikes and LFPs.
• Fabricated the amplifier using a 0.5-mum CMOS process.
• Validated performance through circuit simulations and experimental recordings from avian brain tissue.

Main Results:

• Achieved industry-leading low power consumption (7.56 µW for spikes, 2.08 µW for LFPs).
• Demonstrated excellent noise efficiency factor (NEF) values (2.67 for spikes, 3.21 for LFPs).
• Successfully recorded action potentials from zebra finch brain tissue, validating real-world applicability.

Conclusions:

• The developed neural amplifier represents a significant advancement in energy-efficient neural recording technology.
• The design effectively minimizes noise while maintaining high gain and adjustable bandwidth.
• This amplifier is suitable for advanced brain-computer interfaces and long-term neural monitoring applications.