Effects of Arousal Level and Data-Driven Cardiac Signal Regression on Dynamic Functional Connectivity

Motivation: Resting-state fMRI reveals brain connectivity patterns, but fluctuating arousal can affect dynamic connectivity (DFC). Besides, tracking arousal during fMRI is challenging. Goal(s): Investigate how pupil size, a marker of arousal, modulates DFC during resting-state fMRI. Approach: Participants were classified into high and low arousal groups based on pupil size. Static and dynamic connectivity analyses were performed, using a physiological noise regression method (HAPPY) to account for physiological signals. Results: In the high arousal group, connectivity decreased more in the HAPPY analysis. Dynamic connectivity revealed distinct patterns, with high arousal group showing faster state-transitions, and altered connectivity in the salience and attention networks. Impact: This study demonstrates the importance of tracking arousal levels during resting-state fMRI to better understand brain connectivity dynamics. By using pupil size as an arousal marker, the findings refine rs-fMRI analysis, offering insights into the brain function and physiology.