# It’s all connected! Multivariate pattern analysis of inter-network connectivity distinguishes between reappraisal and passive viewing of emotional scenes

## TL;DR

This study uses brain connectivity patterns to distinguish when people are reappraising emotions versus passively viewing negative images.

## Contribution

The study introduces a network-based MVPA approach to differentiate emotion regulation strategies using inter-network connectivity.

## Key findings

- Reappraisal reduced negative affect compared to passive viewing.
- Inter-network connectivity between ACN and DMN reliably predicted reappraisal versus viewing.
- Univariate analysis showed differences in DMN-ACN connectivity during reappraisal.

## Abstract

Down-regulation using reappraisal is often associated with negative connectivity between prefrontal areas such as the dorsolateral prefrontal cortex (dlPFC) and areas associated with emotion such as the insula and amygdala, though a network perspective is often lacking in emotion regulation research. Whereas the dlPFC is associated with the attentional control network (ACN), the insula and amygdala are associated with the salience and limbic networks, respectively. The default mode network (DMN), including the ventromedial PFC, also contributes to emotion regulation. The present study sought to determine if inter-network functional connectivity can dissociate reappraising from passively viewing a negative image using multivariate pattern analysis (MVPA). Thirty-one participants completed a functional magnetic resonance imaging task in which they reappraised and viewed negative images. Behavioral and skin conductance response results indicated that reappraisal was associated with reductions in negative affect compared to viewing. The univariate connectivity analysis revealed that connections between aspects of the DMN and ACN differed between reappraising versus viewing negative images. Notably, the inter-network connectivity MVPA results demonstrated that whether one was reappraising versus viewing an image could be predicted better than chance, with several connections reliably contributing to the model, including those between ACN and DMN.