Dynamically Allocated Hub in Task-Evoked Network Predicts the Vulnerable Prefrontal Locus for Contextual Memory...

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Systems Neuroscience

Background:

• Prefrontal cortex (PFC) activation varies during memory tasks, but lesion effects differ unpredictably.
• Existing network models based on anatomy don't explain task-specific lesion vulnerability.
• A clear mechanistic link between lesion site and cognitive impairment remains elusive.

Purpose of the Study:

• To test if PFC areas form dynamic, task-specific functional networks centered on a 'functional hub'.
• To determine if lesion-effective sites correspond to these task-specific functional hubs, not static anatomical hubs.
• To provide a mechanistic explanation for differential impairment following PFC lesions.

Main Methods:

• Functional magnetic resonance imaging (fMRI) in macaques performing a temporal contextual memory task.
• Analysis of task-evoked directed connectivity to identify hierarchical, hub-centric networks.
• A novel simulated-lesion method using support vector machines to predict impairment severity.

Main Results:

• Activated PFC areas formed a dynamic, hierarchical network based on task-evoked connectivity, distinct from anatomical networks.
• Simulated lesions predicted impairment severity accurately, matching known dissociations (e.g., area 9/46d vs. 8Ad).
• Lesion effectiveness correlated with network 'hubness' in the functional, task-evoked network, not the anatomical one.

Conclusions:

• Prefrontal cortex dynamically forms functional hub-centric networks for specific cognitive tasks.
• The lesion-effective site is a dynamic functional hub, not a static anatomical one.
• This framework enables precise prediction of behavioral impacts from brain damage or interventions.