A model for attentional information routing through coherence predicts biased competition and multistable perception

Area of Science:

• Neuroscience
• Computational Neuroscience
• Visual Perception

Background:

• Selective attention enables focusing on relevant information and ignoring distractors in visual scenes.
• Neuronal responses in visual area V4 exhibit biased competition, where attended stimuli dominate neuronal activity.
• Information routing is observed through local field potentials correlating with attended stimulus modulation.

Purpose of the Study:

• To propose a coherent framework explaining biased competition and information routing in visual attention.
• To develop and analyze a two-layer spiking cortical network model simulating these attentional mechanisms.

Main Methods:

• A two-layer spiking cortical network model with specific connectivity patterns was constructed.
• The model incorporated distance-dependent lateral connectivity and converging feed-forward connections.
• Network dynamics, including oscillations and phase shifts, were analyzed to understand attentional effects.

Main Results:

• The model successfully reproduced both biased competition and information routing observed in experimental data.
• Lateral interactions and phase shifts between network layers (communication through coherence) were identified as key mechanisms.
• The strength of lateral inhibition was found to be critical in determining network states, transitioning between mixed and bistable representations.

Conclusions:

• The developed spiking neural network model provides a unified explanation for key phenomena of selective visual attention.
• Communication through coherence, mediated by lateral interactions and phase shifts, is highlighted as a fundamental mechanism.
• The model's findings offer insights into multistable perception phenomena like binocular rivalry and the role of lateral inhibition.