Generalized analog thresholding for spike acquisition at ultralow sampling rates

Area of Science:

• Neuroscience
• Signal Processing
• Biomedical Engineering

Background:

• Scaling up multielectrode array (MEA) technology for whole-cortex electrophysiology faces significant data acquisition challenges.
• Standard Nyquist sampling of large MEAs generates massive data volumes, hindering real-time analysis and practical application.
• Existing methods like compressive sensing require initial Nyquist sampling and iterative reconstruction.

Observation:

• Analog thresholding (AT) offers instant spike reconstruction but has limitations in detecting multiple spikes per interval and reconstructing spike width.
• Generalized analog thresholding (gAT) is introduced to overcome limitations of standard AT.
• gAT achieves millisecond temporal resolution at sampling rates as low as 10 Hz.

Findings:

• gAT reduces sampling rates by an order of magnitude compared to AT.
• The method detects multiple spikes per interval and reconstructs spike width.
• gAT provides a simple closed-form, noniterative solution for instant spike reconstruction.
• The technique is robust to hardware nonidealities like quantization error and integration noise.

Implications:

• gAT significantly lowers data acquisition requirements for large-scale electrophysiology.
• The method's hardware-friendly design using integrators and comparators facilitates translation to implantable devices.
• This advancement could enable new scientific investigations and medical technologies for brain activity monitoring.