Face Processing Delays In ASD Are Robust to Variations In Visual Attention

Background:

Face perception is a critical facet of social cognition that develops early in life and plays an enduring and important role in social interaction.  It is subserved by distinct brain mechanisms; event-related potential (ERP) studies reveal an early negative peak (N170) marking initial stages of face processing.  Delays at this stage of face perception have been demonstrated in individuals with ASD, suggesting inefficiency of social perception at its most basic stages. N170 latency is modulated by visual attention to different regions of the face (i.e., eyes evoke a faster N170). Because individuals with ASD show reduced attention to the eyes, it has been suggested that processing delays may reflect differences in visual attention, rather than inefficient processing in face perceptual circuitry.

Objectives:

The study investigated whether face processing delays in ASD reflect inefficient social processing, per se, versus different patterns of attention to the face. We directed visual attention to different parts of the face in typically developing children and children with ASD and analyzed the influence of point of gaze on N170 latency.

Methods:

ERPs were recorded from 13 high-functioning children with ASD and 13 age (~10 years) and sex-matched typical counterparts using a 128 electrode Geodesic Hydrocel Net. In the first condition, point of gaze was manipulated by preceding presentations of neutral faces with fixation points directing attention to the eyes, nose, or mouth; a fourth presentation condition used no fixation crosshair. Attention to intended fixation points was confirmed with a target detection task. In a second condition, we employed a comparable experimental manipulation but integrated eye-tracking (SmartEye Pro v5.5) with EEG recording for a gaze contingent paradigm that presented face stimuli only when participants attended to a fixation point.

Results:

In the first condition, both groups displayed shorter N170 latency when visual attention was directed to the eyes [F(3,72)=38.154; p<.01]; however, irrespective of point of gaze, individuals with ASD displayed longer N170 latency [F(1,24)=7.946; p=.01].  Data analysis for the gaze-contingent condition is in progress, as are analyses relating N170 latency to behavioral measures of social perception. 

Conclusions:

Current results replicate findings of slowed face processing in individuals with ASD.  These delays are not simply byproducts of differences in visual attention; like typical counterparts, children with ASD display faster neural responses when looking at the eyes, but responses are unilaterally delayed irrespective of point of gaze on the face.  This pattern of results is consistent with theories implicating inefficient social information processing as a contributor to the social difficulties experienced by individuals with ASD. Our pending results represent the first integration of concurrent EEG and eye-tracking to create an interactive experimental paradigm responsive to participants’ visual attention.  In addition to providing a more rigorous control of attention, this technical advancement supports development of innovative paradigms for investigating dynamic social interactions.