Implicit Facial Emotion Processing Abilities in Children with ASD

Background:  Social non-verbal behaviour is largely determined by efficient face processing. Individuals with autism spectrum disorder (ASD) are characterized by deficits in social communication and interaction, including difficulties in processing faces. An extensive research tradition suggests that individuals with ASD are less sensitive for socio-communicative information, and pay less attention to the eye region and more to the mouth region.

Objectives:  The present study aims to examine the nature of face processing impairment in 9-to-12 year old children with ASD vs matched typically developing (TD) control children. More specifically, we investigate the subtle and implicit socio-emotional face processing abilities (i.e. facial emotion discrimination and preferential processing of the eye or mouth region) and try to delineate biomarkers that are sensitive at the individual subject level.

Methods:  We use a new innovative approach where we combine scalp electroencephalography (EEG) with fast periodic visual stimulation (FPVS). The general principle of FPVS EEG is that it elicits a steady-state visual evoked potential at exactly the same frequency of visual stimulation. We present images of faces periodically at a 6 Hz base rate and we assess the sensitivity for certain socio-communicative features by periodically entering oddball images displaying changes in expression and/or identity (i.e. every 5th image; 6 Hz/5 = 1.2 Hz oddball rate). Sensitivity for these features can be assessed by quantifying the neural response at the oddball frequency. Participants have to focus on small changes in the fixation cross that is either presented on the eye region or on the mouth region.

Results:  Both TDs and children with ASD display a neural response at the oddball frequency and its harmonics, and responses in both groups are reduced when the faces are inverted. Between group analyses show reduced bilateral occipito-temporal responses in ASD compared to TD. Occipito-temporal responses for both groups are higher when focusing on the mouth as compared to the eyes region.

Conclusions:  With this innovative FPVS EEG method, we studied the implicit emotion discrimination abilities of children with ASD. Preliminary analyses show clear peaks at the oddball frequency and its harmonics in both the ASD and control group. Reduced neural responses in ASD, compared to controls, indicate that TDs are overall better in detecting brief changes in expression when there is no change in identity. However, these results also reveal the ability of children with ASD to implicitly discriminate between facial expressions, suggesting a quantitative difference in emotion processing abilities. Reduced responses in both groups for the inverted images indicate the presence of the inversion effect in ASD and TDs. This highly versatile EEG approach offers an objective and quantifiable index of implicit face processing abilities, reliable at an individual level, within a few minutes of time and without any complex data analyses. These analyses reveal that the technique works robustly in children and clinical populations, and elicits clear peaks at the oddball frequency and its harmonics.