Distributed Phase Oscillatory Excitation Efficiently Produces Attractors Using Spike-Timing-Dependent Plasticity

Area of Science:

• Computational Neuroscience
• Neural Network Modeling
• Brain Information Processing

Background:

• The brain is believed to store information via activity patterns in distributed neuronal groups called attractors.
• Understanding attractor dynamics is crucial for deciphering neural computation.

Purpose of the Study:

• To investigate the efficient creation of distributed attractors in simulated spiking neural networks.
• To explore the formation of associative networks through sequential attractor stimulation.
• To analyze the influence of stimulation parameters on attractor properties.

Main Methods:

• Utilized a randomly connected network of simulated spiking neurons.
• Employed periodic stimulation with distributed phase offsets.
• Applied standard spike-timing-dependent plasticity (STDP) for synaptic modification.
• Analyzed attractor formation, firing patterns, and associative link creation.

Main Results:

• Periodic stimulation combined with STDP efficiently generates distributed attractors.
• Attractors can exhibit ordered or irregular firing patterns based on network conditions.
• Sequential stimulation of attractors, via STDP, forms directed associations, creating an associative network.
• Attractor creation efficiency peaks around 8 Hz for a 20 ms STDP time constant.
• Restimulated attractors exhibit self-oscillation at higher frequencies (10-100 Hz) than the driving frequency.

Conclusions:

• Periodic stimulation and STDP are effective mechanisms for creating neural attractors and associative networks.
• The findings provide insights into potential mechanisms for information representation and learning in the brain.
• The study highlights the role of network parameters, like driving frequency, in shaping neural dynamics.