Real-time covariance computer for cell assemblies is based on neuronal principles

Area of Science:

• Neuroscience
• Computational Neuroscience
• Brain-Computer Interfaces

Background:

• Understanding neural information processing necessitates studying correlated neuronal activities.
• Identifying functional neuron groups requires advanced tools for real-time analysis.
• Existing methods often lack the capability for simultaneous multi-neuron activity assessment.

Purpose of the Study:

• To describe a novel real-time covariance computer designed for analyzing neuronal coupling.
• To present an instrument capable of identifying functional neuron groups based on correlated activity.
• To demonstrate the application of this instrument in analyzing covariant activities in the cat's visual cortex.

Main Methods:

• Development of a real-time covariance computer utilizing leaky integrators to mimic postsynaptic potentials (PSPs).
• Employing analog-to-stochastic converters to simplify covariance calculations via AND gates.
• Utilizing a second set of leaky integrators for deriving individual neuron participation in group activity, with real-time display via LED bargraphs.

Main Results:

• A 4-channel prototype successfully analyzed covariant activities in the cat's visual cortex.
• The instrument effectively calculated neuronal coupling in real time.
• Demonstrated the feasibility of identifying functional neuronal ensembles using the developed system.

Conclusions:

• The developed real-time covariance computer is a viable tool for studying neural information processing.
• The instrument facilitates the identification of functional neuron groups, crucial for understanding brain networks.
• This technology offers a significant advancement in analyzing correlated neuronal activities.