To characterize the brain network dynamics of flexible social cognition in preschoolers with autism spectrum disorder-preschoolers (ASD-p) via analyzing the temporal dynamics of resting-state functional connectivity, and to provide a new perspective for understanding the potential neural mechanisms of their social communications impairment.

Totally 60 ASD-p children(ASD-p group) and 60 matched typical developmental children (TD group) were included from the outpatient department of Affiliated Hospital of Yangzhou University from October to December 2018.Group independent component analysis and sliding window analysis were used to study dynamic functional connectivity based on MATLAB 2013b platform. Finally, k-means cluster analysis was performed to identify distinct functional connectivity states based on the spatial similarity of each functional connectivity pattern, so as to analyze the temporal dynamics of functional connectivity.SPSS 21.0 software was used for data analysis.

Dynamic functional connectivity analysis finally identified three distinct connectivity configurations: an infrequent and more strongly connected state Ⅰ, a more frequent and hypo-connected state Ⅱ, and a relatively frequent and hyper-connected state Ⅲ. The fractional windows of the hypo-connected state Ⅱ in the ASD-p group decreased by 48.04% compared to the TD group, and the fractional windows of states Ⅰ and Ⅲ increased by 22.01% and 32.39%, respectively. The mean dwell time of state Ⅱ in the ASD-p group decreased by 74.89 windows, while the mean dwell time of states Ⅰ and Ⅲ increased by 22.55 windows and 14.84 windows, respectively(Z=-6.080, 6.991, -6.107, all P<0.001, Mann-Whitney U test). The number of transitions in different states of ASD-p was significantly more than that of the TD group ((6.63±3.25)times, (2.65±2.56)times, Z=-6.296, P<0.001, Mann-Whitney U test). Spearman correlation analysis showed that ASD-p social communications impairments was negatively correlated with the mean dwell time of state Ⅱ(r=-0.263, P<0.05), and positively correlated with the number of network state transitions(r=0.272, P<0.05).

The fractional windows and mean dwell time of strongly connected state I and stronger connected state Ⅲ increase proportionally, and the number of transitions increases in ASD-p patients. Dynamic functional connectivity characteristics related to the core symptoms of ASD-p may be used as a reliable neuroimaging biomarker.

Autism spectrum disorder; Preschoolers; Social communications dysfunction; Dynamic function connectivity