Activity flow under the manipulation of cognitive load and training

Area of Science:

• Neuroscience
• Cognitive Science
• Computational Neuroscience

Background:

• Flexible cognitive functions, like working memory (WM), depend on balancing localized and distributed brain processing.
• Understanding the specific contributions of local versus distributed processing to task-induced brain activity remains a challenge.
• Activity flow mapping offers a way to assess distributed processing, but its role in adaptive cognitive changes is underexplored.

Purpose of the Study:

• To investigate how working memory load and adaptive training modulate activity flow and brain activation in frontoparietal systems.
• To differentiate the contributions of within-network versus between-network activity flow to cognitive processing.
• To establish a methodological framework for studying cognitive plasticity and information integration.

Main Methods:

• Recruited 51 healthy volunteers (31 female) for a working memory task with varying load conditions and adaptive training.
• Utilized activity flow mapping to analyze brain activation and connectivity patterns in the frontoparietal network.
• Examined changes in within-network and between-network activity flow before and after training.

Main Results:

• At baseline, both executive control network (ECN) and dorsal attention network (DAN) activation increased with WM load; only DAN showed a linear distributed processing response (within-network).
• Adaptive training increased distributed processing in the ECN and induced a linear load response, primarily via between-network activity flow.
• Activity flow prediction demonstrated a causal link between training, connectivity, and brain activity.

Conclusions:

• Activity flow mapping provides unique insights into neural processing beyond traditional activation measures, especially under cognitive load and training.
• Adaptive training enhances distributed processing in the ECN through increased between-network communication.
• This study introduces a novel methodological approach for investigating information integration and segregation in cognitive systems.