Modularity-dependent storage of dynamic spiking patterns: Bridging micro- and mesoscopic representations.

Area of Science:

• Computational neuroscience
• Neural network modeling
• Cognitive dynamics

Background:

• Biological systems, especially the brain, utilize asynchronous and overlapping dynamics for concurrent processes.
• Cognitive tasks involve distributed, interacting brain regions, not isolated sequential ones.

Purpose of the Study:

• Investigate mechanisms of neural coordination in complex biological systems.
• Model how temporal overlap in neural activity influences information processing and memory.

Main Methods:

• Simulated a modular spiking neural network with leaky integrate-and-fire neurons.
• Incorporated spike-timing-dependent plasticity and a parameter (η) for temporal overlap.
• Analyzed network dynamics across sequential to overlapping activation regimes.

Main Results:

• The model stores spatiotemporal patterns at mesoscopic and microscopic levels.
• Tuning temporal overlap (η) shifted network dynamics from sequential to overlapping regimes.
• Network organization and spatiotemporal coding impact memory capacity and retrieval.

Conclusions:

• Temporal structure and network organization are crucial for robust memory storage and replay.
• Overlapping dynamics support efficient encoding and retrieval of multiple patterns.
• Findings provide insight into large-scale brain function and memory mechanisms.