To study the differences in functional brain networks between eyes-closed (EC) and eyes-open (EO) at resting state, electroencephalographic (EEG) activity was recorded in 21 normal adults during EC and EO states. The synchronization likelihood (SL) was applied to measure correlations between all pairwise EEG channels, and then the SL matrices were converted to graphs by thresholding. Graphs were measured by topological parameters in theta (4–7 Hz), alpha (8–13 Hz), and beta (14–30 Hz) bands. By changing from EC to EO states, mean cluster coefficients decreased in both theta and alpha bands, but mean shortest path lengths became shorter only in the alpha band. In addition, local efficiencies decreased in both theta and alpha bands, while global efficiencies in the alpha band increased inversely. Opening the eyes decreased both nodes and connections in frontal area in the theta band, and also decreased those in bilateral posterior areas in the alpha band. These results suggested that a combination of the SL and graph theory methods may be a useful tool for distinguishing states of EC and EO. The differences in functional connectivity between EC and EO states may reflect the difference of information communication in human brain.

fMRI studies have suggested that there are two different attention networks: the dorsal attention network (DAN) and the ventral attention network (VAN). The emotion network has also been discovered by some researches. The dorsal attention network controls goal-oriented top-down deployment of attention; the ventral attention network controls stimulus-driven bottom-up deployment of attention; the emotion network will feed back the stimulus especially fearful expressions from the environment. The interaction within these networks has been noticed but few studies have been carried out. The purpose of this study is to explore the interaction within these networks. The regions of interest were acquired by using the GLM analysis after which the granger causality among these ROIs was calculated. Connections among ROIs were considered as causal when their respective granger causality value is greater than the mean value of all granger causalities. According to the results, there is interaction within the three networks, which suggested that the ventral attention network may distract the dorsal attention network and that the emotion network may influence both attention networks.

We examined the influence of different emotional cues (happy or sad) on temporal attention (short or long interval) using behavioral as well as event-related potential recordings during a Stroop task. Emotional stimuli cued short and long time intervals, inducing 'sad-short', 'sad-long', 'happy-short', and 'happy-long' conditions. Following the intervals, participants performed a numeric Stroop task. Behavioral results showed the temporal attention effects in the sad-long, happy-long, and happy-short conditions, in which valid cues quickened the reaction times, but not in the sad-short condition. N2 event-related potential components showed sad cues to have decreased activity for short intervals compared with long intervals, whereas happy cues did not. Taken together, these findings provide evidence for different modulation of sad and happy facial cues on temporal attention. Furthermore, sad cues inhibit temporal attention, resulting in longer reaction time and decreased neural activity in the short interval by diverting more attentional resources.

We examined the influence of spatial selective attention on the processing of emotional faces (happy neutral, and fear) using behavioral as well as event-related potential recordings. Emotional stimuli were rapidly presented randomly to the right or the left visual field while participants attended to one visual field at a time, detecting smaller stimuli that were shown in the attended field. Behavioral results showed decreased accuracy for the fearful faces compared with neutral and happy faces. Event-related potential data showed that compared with the neutral and happy faces, fearful faces appearing in the right visual field elicited enhanced contralateral P1 amplitudes in the unattended condition, whereas fearful faces appearing in the left visual field elicited decreased contralateral N170 activity in the attended condition. These findings provide evidence for differential emotional processing under spatial attended and unattended conditions.