Evaluating performance of neural codes in model neural communication networks

Area of Science:

• Computational Neuroscience
• Systems Neuroscience
• Neural Coding

Background:

• Neurons utilize diverse spatio-temporal encoding strategies to transmit information.
• Understanding the efficiency and suitability of different neural codes under varying network conditions remains a significant challenge in neuroscience.

Purpose of the Study:

• To model small networks of Hindmarsh-Rose neurons to investigate the performance of different neural codes.
• To quantify the rate of information exchange using temporal and firing-rate codes.

Main Methods:

• Simulated small networks of chemically and electrically coupled Hindmarsh-Rose spiking neurons.
• Focused on membrane potentials and phases to analyze temporal and firing-rate codes.
• Numerically estimated the Mutual Information Rate to quantify information exchange.

Main Results:

• Firing-rate and interspike-interval codes demonstrate greater robustness against additive Gaussian white noise.
• In small, noise-free networks, temporal codes (spike-timing, phase) facilitate high information exchange between adjacent neurons.
• Non-adjacent neurons show higher information exchange rates with firing-rate and interspike-interval codes.

Conclusions:

• Temporal codes may be favored in small neural microcircuits for direct information exchange.
• Firing-rate and interspike-interval codes could be more efficient for information transfer across larger, sparser neural networks.
• The choice of neural code's efficiency depends on network connectivity and scale.