There is a substantial body of evidence from both electrophysiological and neuroimaging studies showing face processing impairments in adults and children with autism. Most of this research has focused on high functioning subjects with very few studies carried out on low functioning samples and still fewer studies assessing preschoolers. Additionally, very few studies have examined the neurophysiological changes in face processing as a result of intervention.
Objectives:
The goal of the present study is to fill this gap in the literature by assessing brain changes in face processing associated with treatment. We assessed the N170 event-related potential (ERP) from faces to assess the change in N170 amplitude, latency and cortical location after one year of intervention using DIR/floortime (Developmental Individualized Relationship model developed by Greensan and Wieder, 2007). Given that previous research has identified smaller N170 amplitude ERPs, longer N170 latencies and source activations in areas outside of the fusiform gyrus (e.g., inferior temporal gyrus and middle temporal gyrus), it was hypothesized that successful treatment would be related to increased N170 amplitude, shorter component latencies and a shift in activation from compensatory regions to activation in the region of the fusiform gyrus.
Methods:
Dense-array (128 channel - EGI) data was collected from twenty-three, two to five year old children who met criteria for a diagnosis of autism based on ADOS and ADI and twenty-six typically developing age-matched controls. Participants were provided with an EEG net desensitization training protocol until they tolerated wearing the net for 30 minutes. Face stimuli were presented using Eprime (PST). An eye-tracker (Tobii) integrated in real-time with the EEG data stream was used to ensure that trials contained correct face gaze at stimulus onset. Faces were presented for 1500 milliseconds. Data were analyzed using Netstation (EGI) and inverse modeling was carried out using Loretta in Geosource (EGI). Behavioural changes associated with treatment success in the domain of social interaction were assessed using the Child Behaviour Rating Scale (Mahoney). Behavioural and electrophysiological assessments were carried out before and after 12 months of intervention using the DIR/Floortime method.
Results: The N170 showed a significant increase in amplitude and a decrease in component latency across the one year treatment period. Source estimation for the N170 revealed a shift in activation in the clinical group from middle temporal gyrus and inferior temporal gyrus at pre-treatment to greater activation in the area of the right fusiform gyrus at post-treatment. Controls showed a pattern of activation similar across time points with greatest activation in the region of the fusiform gyrus. Behavioural results mirrored changes in ERP activity with a significant increase in social involvement, attention to social activity, initiation of joint attention and overall enjoyment of interaction.
Conclusions:
Results from this study provide support for the proposal that successful treatment in young children with autism is related to changes in areas of the brain associated with social information processing.