Altered dynamic network interactions in children with ASD during face recognition revealed by time-varying EEG networks

Abstract

Abstract Background: Although the electrophysiological event-related potential (ERP) in face processing (e.g., N170) is widely accepted as a face-sensitivity biomarker that is deficient in children with autism spectrum disorders (ASD), the time-varying brain networks during face recognition are still awaiting further investigation.Methods: To explore the social deficits in ASD, especially the time-varying brain networks during face recognition, the current study analyzed the N170, cortical activity, and time-varying networks under three tasks (face-upright, face-inverted, and house-upright) in ASD and typically developing (TD) children.Results: The results revealed a smaller N170 amplitude in ASD compared with TD, along with decreased cortical activity mainly in occipitotemporal areas. Concerning the time-varying networks, the atypically stronger information flow and brain network connections across frontal, parietal, and temporal regions in ASD were reported, which reveals greater effort was exerted by ASD to obtain comparable performance to the TD children, although the amplitude of N170 was still smaller than that of the TD children.Conclusion: Different brain activation states and interaction patterns of brain regions during face-processing were discovered between ASD and TD. These findings shed light on the face-processing mechanisms in children with ASD and provide new insight for understanding the social dysfunction of ASD.