Inter-subject phase synchronization differentiates neural networks underlying physical pain empathy

Abstract

Recent approaches for understanding the neural basis of pain empathy emphasize the dynamic construction of networks underlying this multifaceted social cognitive process. Inter-subject phase synchronization (ISPS) is an approach for exploratory analysis of task-fMRI data that reveals brain networks dynamically synchronized to task-features across participants. We applied ISPS to task-fMRI data assessing vicarious pain empathy in healthy participants (n = 238). The task employed physical (limb) and affective (face) painful and corresponding non-painful visual stimuli. ISPS revealed two distinct networks synchronized during physical pain observation, one encompassing anterior insula and midcingulate regions strongly engaged in (vicarious) pain and another encompassing parietal and inferior frontal regions associated with social cognitive processes which may modulate and support the physical pain empathic response. No robust network synchronization was observed for affective pain, possibly reflecting high inter-individual variation in response to socially transmitted pain experiences. ISPS also revealed networks related to task onset or general processing of physical (limb) or affective (face) stimuli which encompassed networks engaged in object manipulation or face processing, respectively. Together, the ISPS approach permits segregation of networks engaged in different psychological processes, providing additional insight into shared neural mechanisms of empathy for physical pain, but not affective pain, across individuals.

Keywords: affective pain; brain synchronization; inter-subject correlation; pain empathy; physical pain.

Fig. 1

Activation maps (unthresholded) from the conventional GLM approach. The BOLD activation maps for each task condition (middle) as compared to baseline (left) or the relative subtraction contrasts between pain and control in physical and affective conditions, respectively (right).

Fig. 2

Synchronization of the nine components. (C = component; R = right; L = left). The left panel displays correlation coefficients of each component and the most associated task condition. The middle panel displays the synchronization time courses for each component and the GLM reference functions it is most associated with. The right panel shows the spatial weights for each component [visualized with BrainNet Viewer (Xia et al., 2013)].