Spatiotemporal dynamics of working memory under the influence of emotions based on EEG

J Neural Eng. 2020 Apr 29;17(2):026039. doi: 10.1088/1741-2552/ab7f50.

Abstract

Objective: Previous studies have reported that working memory (WM) may be affected by emotions and that the effect may exist in different stages of WM. However, at present it remains controversial whether emotions inhibit or facilitate WM, and how the mechanism of dynamic information transmission in the brain during WM is affected by emotions.

Approach: In this study, we used a video database to induce three emotions (negative, neutral, and positive) and adopted a change detection paradigm based on electroencephalography. Event-related potential (ERP) analysis, event-related spectral perturbation analysis, source location analysis based on the dipole localization method and the distributed source localization method, and effective connectivity analysis were performed.

Main results: Both behavioral and ERP results suggest that positive emotions have no significant effect on WM capacity, while negative emotions could facilitate WM capacity. Furthermore, the effective connectivity results based on two source location methods suggest that the long-range connectivity between the frontal and posterior areas can reflect the influence of positive and negative emotions on the WM network, in which the connectivity under the positive emotion condition occurs in the earlier period of WM maintenance, while the connectivity under the negative emotion condition occurs in the later period of WM maintenance.

Significance: The consistency of the behavioral, ERP, and effective connectivity results suggests that under the negative emotion condition, the top-down attention modulation between the frontoparietal area and posterior area could promote the most relevant information storage during WM maintenance.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Electroencephalography*
  • Emotions
  • Evoked Potentials
  • Frontal Lobe
  • Memory, Short-Term*