Event-driven neural integration and synchronicity in analog VLSI.

Area of Science:

• Neuroscience
• Computational Neuroscience
• Neuromorphic Engineering

Background:

• Neural synchrony and temporal coding are crucial for brain function, generating local field potentials and complex dynamics.
• Spike event timing enables coincidence detection, offering a robust neural code resistant to noise.

Purpose of the Study:

• To present a novel spike-based coincidence detection system.
• To implement this system using integrate-and-fire neural and synaptic dynamics within a hierarchical architecture.

Main Methods:

• Utilized integrate-and-fire neural membrane dynamics and pooled conductance-based synaptic dynamics.
• Developed a hierarchical address-event architecture to encode synaptic events with connectivity, strength, and delay parameters.
• Configured global parameters for neural and synaptic temporal dynamics.

Main Results:

• Successfully demonstrated spike-based coincidence detection within the implemented architecture.
• Analyzed the observed coincidence detection through measurements on a log-domain analog VLSI chip.
• Validated the robustness of the neural code against incoherent noise.

Conclusions:

• The proposed architecture effectively implements spike-based coincidence detection.
• Analog VLSI implementation confirms the feasibility of this neural coding strategy.
• This approach offers a promising avenue for understanding neural computation and developing neuromorphic systems.