A single-chip signal processing and telemetry engine for an implantable 96-channel neural data acquisition system

Area of Science:

• Neuroscience
• Biomedical Engineering
• Electrical Engineering

Background:

• Clinical viability of cortical brain-machine interfaces (BMIs) relies on fully implantable neural data acquisition systems.
• Efficient processing of multi-unit neural data is crucial for real-time BMI operation.
• Bidirectional wireless communication is essential for untethered BMI functionality.

Purpose of the Study:

• To design and implement a single-chip device for processing neural data in a fully implantable BMI system.
• To address the digital processing needs for 96 channels of neural data and wireless communication.
• To provide a versatile platform for various data output modes suitable for BMI applications.

Main Methods:

• Implementation of a single programmable logic device (PLD) for neural data processing.
• Integration of data reduction algorithms for 96 channels of multi-unit neural data.
• Development of a bidirectional telemetry interface for wireless transceiver communication and command interpretation.

Main Results:

• The device processes neural data sampled at 31.25 kHz, outputting a line-encoded serial bitstream.
• Supports four data output modes: single-channel streaming, extracted spike waveforms, bincounts, or combined streaming and spikes.
• Capable of outputting up to 2000 extracted spikes per second with low latency suitable for BMIs.

Conclusions:

• The developed single-chip device provides essential digital processing for fully implantable neural data acquisition systems.
• This implementation facilitates the advancement of clinically viable cortical brain-machine interfaces.
• The system's flexibility in data handling and wireless communication capabilities supports diverse BMI research and applications.