Energy Consumption Optimization, Response Time Differences and Indicators in Cortical Working Memory Revealed by Nonequilibrium

TL;DR

This paper introduces a nonequilibrium landscape flux approach to analyze cortical networks, revealing energy costs, temporal hierarchies, and state transitions in brain dynamics, with implications for understanding working memory and cognition.

Contribution

It presents a novel theoretical framework using nonequilibrium landscape flux analysis to quantify cortical dynamics, energy consumption, and hierarchical processing in large-scale brain networks.

Findings

Quantifies energy costs associated with cortical network activity.

Identifies temporal hierarchies linked to stimuli distribution.

Provides tools for predicting cortical state transitions.

Abstract

The neocortex, a complex system driving multi-region interactions, remains a core puzzle in neuroscience. Despite quantitative insights across brain scales, understanding the mechanisms underlying neural activities is challenging. Advances from Hopfield networks to large-scale cortical models have deepened neural network theory, yet these models often fall short of capturing global brain functions. In large-scale cortical networks, an intriguing hierarchy of timescales reflects diverse information processing speeds across spatial regions. As a non-equilibrium system, the brain incurs significant energy costs, with long-distance connectivity suggesting an evolutionary spatial organization. To explore these complexities, we introduce a nonequilibrium landscape flux approach to analyze cortical networks. This allows us to quantify potential landscapes and principal transition paths, uncovering dynamical characteristics across timescales. We examine whether temporal hierarchies correlate with stimuli distribution and how hierarchical networks exhibit differential responses. Furthermore, our analysis quantifies the thermodynamic cost of sustaining cognition, highlighting a link to network connectivity. These findings provide insights into energy consumption during cognitive processes and emphasize the spatial benefits for working memory tasks. Experimental validation is challenging due to evolutionary variability, making our theoretical approach valuable for quantifying complex dynamics. By assessing time irreversibility and critical slowdown, we gain predictive insights into network bifurcations and state transitions, offering practical tools for identifying cortical state changes. These results advance our understanding of cortical dynamics.