Edge-Based General Linear Models Capture Moment-to-Moment Fluctuations in Attention

Area of Science:

• Neuroscience
• Cognitive Neuroscience
• Functional Neuroimaging

Background:

• Sustaining attention is crucial but fluctuates.
• Functional connectivity (FC) networks predict attention, but traditional methods lack temporal precision.
• Dynamic FC analysis needs methods to capture rapid, moment-to-moment network changes.

Purpose of the Study:

• To apply novel edge time series analysis to fMRI data.
• To capture rapid, moment-to-moment fluctuations in brain networks related to attention.
• To investigate event-based and parametric changes in functional connectivity.

Main Methods:

• Utilized "edge cofluctuation time series" to analyze timepoint-by-timepoint region cofluctuations.
• Applied event-based and parametric fMRI analyses to edge time series.
• Examined two independent fMRI datasets of young adults performing a sustained attention task.

Main Results:

• Identified specific "edges" (connections) that rapidly change with rare task events.
• Found other edges that correlate with continuous fluctuations in attention.
• Demonstrated that edge-based changes are not fully explained by univariate activity patterns.

Conclusions:

• Edge time series analysis provides high temporal precision for dynamic FC.
• This approach reveals rapid network reconfigurations critical for attention.
• Combines traditional fMRI with edge-based methods for deeper insights into brain dynamics.