Cerebral coherence in task-based fMRI hyperscanning: a new trick to an old dog

This study features an fMRI hyperscanning experiment, mapping the brains of the dyads from two fMRI sites, 305 km apart. There are two conditions: in half of the trials (the cooperation condition), the dyad had to collaborate to win and then split the reward, whereas in the other half (the competition condition), the winner took all the reward, thereby resulting in dynamic strategic interactions. Each subject took alternating turns as senders and receivers. To calculate the cerebral coherence in such jittered event-related fMRI tasks, we first estimated the feedback-related BOLD responses of each trial, using 8 finite impulse response functions (16 seconds), and then concatenated the beta volume series. With the right temporal-parietal junction (rTPJ) as the seed, the interpersonal connected brain areas in the cooperation and competition conditions were separately identified: the former condition with the right superior temporal gyrus (rSTG) and the latter with the left precuneus (lPrecuneus) (as well as some other regions of interest), both peaking at the designated frequency bin (1/16 s = 0.0625 Hz), but not in permuted pairs. In addition, the extended coherence analyses on shorter (12 s, or .083 Hz) and longer (20 s, or .05 Hz) concatenated volumes verified that only approximately in the trial length were the rTPJ-rSTG and rTPJ-lPrecuneus couplings found. In sum, our approach both showcases a flexible analysis method that widens the applicability of interpersonal coherence in the rapid event-related fMRI hyperscanning, and reveals a context-based interpersonal coupling between pairs in cooperation vs. competition. Author summarySocial neuroscience is gaining momentum, while coherence analysis as one of the interpersonal connectivity measures is rarely applied to the rapid event-related fMRI. The reason could be that the inherent task design (such as the periodicity constraint for Fourier transformation), among others, limits its applicability and usage. In this fMRI hyperscanning study of a two-person strategic interactions, we independently estimated the feedback-related BOLD responses of each trial, and concatenated the beta time series for interpersonal coherence. The main advantage of this method is in its flexibility against the constraints of jittered experimental trials intermixing several task conditions in most task-based fMRI runs. In addition, our coherence results, which highlight two inter-brain couplings (e.g., rTPJ-rSTG between collaborating, and rTPJ-lPrecuneus for competing dyads) among other brain regions, plus its temporal specificity of such seed-brain couplings only between pairs, both replicate previous run-wide fMRI coherence results, and hold great promise in extending its applicability in task-based fMRI hyperscanning.