Information coding and oscillatory activity in synfire neural networks with and without inhibitory coupling

Area of Science:

• Computational neuroscience
• Neural network dynamics
• Complex systems analysis

Background:

• Neural populations exhibit synchronous activity that can either stabilize or decay.
• Understanding how input intensity and network structure influence this activity is crucial.

Purpose of the Study:

• To investigate the impact of noise on neural population activity.
• To explore the role of inhibitory coupling in generating complex firing patterns.
• To analyze neural network dynamics using a Markov chain model.

Main Methods:

• Analysis of pulse-packet propagation as a Markov chain.
• Numerical simulations using integrate-and-fire and compartmental neuron models.
• Eigenvalue and eigenvector analysis of the transition matrix.

Main Results:

• Pulse-packet survival probability (firing rate) correlates with input intensity in the presence of noise.
• Inhibitory coupling leads to quasi-periodic alternation between firing activity levels.
• Simulation results align with Markov model predictions, showing pool activation depends on preceding pool activity.

Conclusions:

• Noise-driven neural networks exhibit input-dependent activity survival.
• Inhibitory interactions can induce complex oscillatory dynamics in neural networks.
• The Markov chain model provides a robust framework for understanding neural population dynamics.