Multichannel spike detector with an adaptive threshold based on a Sigma-delta control loop.

Area of Science:

• Neuroscience
• Signal Processing
• Embedded Systems Engineering

Background:

• Electrophysiological recordings generate large datasets requiring efficient spike detection.
• Real-time processing of multi-channel neural data presents significant computational challenges.
• Existing spike detection methods may struggle with low signal-to-noise ratios (SNRs) and high power consumption.

Purpose of the Study:

• To develop and validate a novel digital spike detector for parallel processing of 32 electrophysiological channels.
• To implement an adaptive thresholding algorithm optimized for real-time performance and low power consumption.
• To evaluate the performance of the proposed spike detector against established off-line software.

Main Methods:

• A Sigma-delta control loop was employed to estimate signal noise and optimize detection.
• A robust algorithm, independent of input signal amplitude, was utilized for adaptive thresholding.
• The spike detector was implemented on a Spartan-6 Field-Programmable Gate Array (FPGA) using basic logic blocks and a low clock frequency (<6 MHz).

Main Results:

• The system achieved 100% true positive detection rate for SNRs down to 5 dB.
• A true positive detection rate of 62.3% was achieved at an SNR as low as -2 dB with a firing rate of 150 action potentials per second (AP/s).
• The FPGA implementation demonstrated competitive performance compared to dedicated off-line spike detection software, utilizing minimal resources and low power.

Conclusions:

• The proposed digital spike detector offers an efficient, low-power solution for real-time, multi-channel electrophysiological data analysis.
• The adaptive thresholding scheme provides robust spike detection across a wide range of SNRs.
• This system is suitable for portable and resource-constrained neuroscientific research applications.