Exploring the Relationship between Visual Evoked Potentials and Visual Search in ASD: Results from the ABC-CT Interim Analysis

Background: Visual Evoked Potentials (VEP) are electrophysiological signals extracted from the visual cortex that are activated when visual pattern changes occur quickly (Odom et al., 2009). Research suggest that adolescent and adults with Autism Spectrum Disorder (ASD) exhibit lower P100 amplitudes in their VEPs compared to the Typically Developing (TD) individuals (Kovarski et al., 2016). It remains unknown how, or if these differences in brain activity are associated with corresponding higher order mechanisms of visual sampling. Visual pathways carry information to areas of the brainstem that organize and effect eye movements. Thus, the timing and organization of early sensory response may affect the planning and organization of eye-movements at a more cognitive level. Consequently, atypical low-level visual processing may be reflected in atypical Visual Search (VS) behavior, as exemplified by eye-tracking literature that has shown less exploratory eye scanning in children with ASD (Sasson et al., 2008).

Objectives: To explore the relationship between low level perceptual visual processing as measured by VEP and potentially higher-level visual search behavior as measured by the proportion of time spent in fixations in a VS paradigm in children with and without ASD.

Methods: 6-11 year olds with ASD (n=126) and TD (n=55) watched videos of flickering checkerboards (VEP) and viewed static image arrays, with each array composed of five social and nonsocial images (VS). Low-level visual processing was measured through P100 amplitudes during VEP. Higher-level visual processing was measured by fixation ratios (total fixation duration/valid looking time). Fixations were calculated by Li et al. (2016) with 1° spatial threshold and 100ms minimum fixation time. Correlations and ANCOVAS examined the relationships between the variables.

Results: TD showed higher VS fixation ratios (p=.009), suggesting that TD children spent more time extracting information from the scene than the ASD group, and controlling for IQ did not alter this relationship. Consistently, in the ASD group, higher VS fixation ratios were associated with less social impairment (higher ADOS Social Affect severity scores) (r=.19, p=.037). There were no between-group differences in P100 amplitude (p>.05), but within the TD group lower P100 amplitudes were associated with greater social impairment (r=-.15, p=.046). There were no relationships between P100 amplitude and fixation ratios within any of the groups or when the groups were combined even when controlling for age, ADOS, and/or IQ scores.

Conclusions: P100 amplitudes and eye tracking fixation measures were not found to be related in this study. However the VS results suggest that higher-order visual information processing may be impaired in ASD, with reduced visual sampling being associated with more autism symptomatology. The VEP results suggest that, while between-group differences were not observed, it is capturing some aspect of the continuum of social function across the TD population. The lack of relationship between the VEP and VS variables might be due to task differences between the EEG and ET paradigms, but could also signal a fundamental disconnect between neural and neurobehavioral markers. Lower-level visual perception may not prominently influence the higher-order visual exploratory deficits evident in children with ASD.