Identifying EEG Responses Modulated by Working Memory Loads from Weighted Phase Lag Index Based Functional Connectivity Microstates

Abstract

Working-memory training has been viewed as an important intervention way to improve the working memory capacity of children’s brain. However, effective electroencephalogram (EEG) features and channel sites correlated with working memory loads still need to be identified for future application to brain-computer interface (BCI) system. In this experiment, 21 adolescent subjects’ EEG was recorded while they performed an n-back working-memory task with adjustable loads (n = 1, 2, 3). Based on global neuronal workspace (GNW) theory, α-band (4–8 Hz) weighted phase lag index (wPLI) between signals was computed in consecutive 200-ms time windows of each trial to construct continuously evolving functional connectivity microstates. Statistical analysis reveals that, in post-stimulus 200–400 ms and 400–600 time intervals, working-memory loads significant modulate functional integration of global network, showing increasing connectivity density and decreasing characteristic path length with the increase of memory loads. Classifications between single-trail samples from high- and low-loads were conducted for local nodal connection strength. Analytical results indicate that network vertices in right-lateral prefrontal cortex, right inferior frontal gyrus and pre-central cortices are highly involved in identifiable brain responses modulated by working-memory loads, suggesting feasible EEG reference locations and novel features for future BCI study on the development of children/adolescents’ working memory resource.