Connectome-based Models Predict Separable Components of Attention in Novel Individuals

Area of Science:

• Neuroscience
• Cognitive Science
• Brain Imaging

Background:

• Attention is often viewed as a unified process, but it comprises multiple independent components.
• Understanding the neural basis of these distinct attentional functions is crucial for cognitive neuroscience.

Purpose of the Study:

• To investigate if key components of attention, as defined by Posner and Petersen's model (alerting, orienting, executive control), are reflected in the brain's intrinsic functional organization.
• To apply connectome-based predictive modeling (CPM) to predict these attention components from functional connectivity data.

Main Methods:

• Participants underwent functional magnetic resonance imaging (fMRI) while performing the Attention Network Task (ANT) and during rest.
• Connectome-based predictive modeling (CPM) with leave-one-subject-out cross-validation was used to predict ANT performance metrics from functional connectivity.
• Models were tested using both task-based and resting-state fMRI data.

Main Results:

• CPM successfully predicted overall ANT accuracy, reaction time (RT) variability, and executive control scores from task-based functional connectivity.
• Alerting scores were predicted from resting-state functional connectivity alone, indicating task-independent neural signatures.
• A sustained attention model also predicted performance metrics, suggesting overlap between executive control and sustained attention.

Conclusions:

• The brain's functional organization reflects distinct components of attention.
• The neural infrastructure supporting alerting is separable from other attention networks and measurable during rest.
• CPM is a valuable tool for dissecting attention's components and their neural underpinnings.