Multi-stability and pattern-selection in oscillatory networks with fast inhibition and electrical synapses

Area of Science:

• Computational neuroscience
• Network dynamics
• Systems biology

Background:

• Oscillatory neural networks can exhibit multiple stable activity patterns.
• Switching between these patterns is induced by transient signals, but signal properties are poorly understood.

Purpose of the Study:

• Investigate signal properties that induce switching between in-phase (IP) and anti-phase (AP) activity patterns in neural networks.
• Explore how stimulus profiles encode information in network dynamics.

Main Methods:

• Numerical simulations of neural networks of varying sizes (2, 4, 6 cells) composed of relaxation oscillators.
• Analysis of susceptibility windows for pattern switching.
• Application of depolarizing and hyperpolarizing stimuli to specific cell subpopulations.

Main Results:

• Susceptibility windows for switching between IP and AP patterns are consistent across different network sizes.
• Specific stimulus profiles can evoke switching to different AP patterns.
• The resulting network activity pattern encodes the profile of the applied stimulus.

Conclusions:

• Transient signals with specific profiles can reliably induce pattern switching in oscillatory networks.
• The network's response encodes information about the stimulus, suggesting a mechanism for information storage.
• These findings apply to various network architectures, including those with electrical synapses and cross-inhibition.