Altered Dynamic Functional Network Connectivity in Healthy Adults with Acute Pain: Findings from the Human Connectome Project

Characterizing the neural signature of pain and its modulation is critical for assessing treatment efficacy and conducting translational clinical research. However, the dynamics of pain processing in the brain have remained largely unknown. In this study, we employed independent component analysis (ICA) as a data-driven clustering method on resting-state functional magnetic resonance imaging (fMRI) to obtain intrinsic connectivity networks (ICNs) in a cohort of healthy adults from the Human Connectome Project (HCP) who were identified as having acute pain. We examined the temporal dynamic functional network connectivity (dFNC) with sliding time window correlation and k-means clustering, and compared dFNC state properties and meta-state metrics between groups. Results showed that acute pain had a significant impact on dFNC in a common connectivity state (dynamic state 5) among several ICN pairs, including the salience network, default mode network, central executive, dorsal attention networks, and basal ganglia (false discovery rate [FDR]-corrected p of 0.05). Furthermore, healthy adults with and without acute pain exhibited differences in mean dwell time (dynamic state 3), which indicated that individuals with acute pain spent more time in particular states than those without pain. Meta-state dynamic analysis further indicated significant group differences in the number of states (i.e., unique time windows for each subject), changes between states (i.e., number of times each subject changes from one meta-state to other), and total travelled distances. These preliminary results provide new information about time-varying properties of pain states related to acute pain and advocate for further state-based analyses of pain for future pain biomarker discovery and development.