The neural bases of resilient semantic system: evidence of variable neuro-displacement in cognitive systems

Je Young Jung,Grace E Rice,Matthew A Lambon Ralph

doi:10.1007/s00429-021-02272-1

Je Young Jung, Grace E Rice + Show 1 more

Open Access

https://doi.org/10.1007/s00429-021-02272-1

Copy DOI

Abstract

The purpose of this study was to explore an important research goal in cognitive and clinical neuroscience: What are the neurocomputational mechanisms that make cognitive systems “well engineered” and thus resilient across a range of performance demands and to mild levels of perturbation or even damage? A new hypothesis called ‘variable neuro-displacement’ suggests that cognitive systems are formed with dynamic, spare processing capacity, which balances energy consumption against performance requirements and can be resilient to changes in performance demands. Here, we tested this hypothesis by investigating the neural dynamics of the semantic system by manipulating performance demand. The performance demand was manipulated with two levels of task difficulty (easy vs. hard) in two different ways (stimulus type and response timing). We found that the demanding semantic processing increased regional activity in both the domain-specific semantic representational system (anterior temporal lobe) and the parallel executive control networks (prefrontal, posterior temporal, and parietal regions). Functional connectivity between these regions was also increased during demanding semantic processing and these increases were related to better semantic task performance. Our results suggest that semantic cognition is made resilient by flexible, dynamic changes including increased regional activity and functional connectivity across both domain-specific and domain-general systems. It reveals the intrinsic resilience-related mechanisms of semantic cognition, mimicking alterations caused by perturbation or brain damage. Our findings provide a strong implication that the intrinsic mechanisms of a well-engineered semantic system might be attributed to the compensatory functional alterations in the impaired brain.

Highlights

One critical feature of any well-engineered system is its resilience to variable performance demands, perturbation and minor damage
In the response timing manipulation (RM) dataset, the accuracy was significantly decreased in the hard condition relative to the easy condition, whereas the reaction time (RT) was faster in the hard condition due to the response time manipulation (p < 0.001)
Beyond investigations of executive function, other domains of higher cognition are often examined at a single performance demand level in each study whilst other parameters of interest are manipulated

Summary

Introduction

One critical feature of any well-engineered system is its resilience to variable performance demands, perturbation and minor damage. How resilience is achieved in higher cognitive systems is important both for cognitive and clinical neuroscience. Rarely considered in laboratorybased explorations of higher cognition (beyond executive function where demand variations are inherently important), in everyday life we are faced with and are resilient to variations in task difficulty, degraded stimuli, etc. A core tenet in engineering is to design a system that is resilient to functional stresses, as well as to balance performance and energy costs. Under standard levels of performance demand the full neural system will be downregulated to save energy but have spare capacity that can be utilised when the situation necessitates it Under standard levels of performance demand the full neural system will be downregulated to save energy but have spare capacity that can be utilised when the situation necessitates it (Stefaniak et al. Vol.:(0123456789)

Objectives

Methods

Results

Conclusion