Eye Tracking Utilizing Age Matched Social Scenes and Geometric Shapes

Background: Social impairment is a key characteristic of patients with Autism Spectrum Disorders (ASD). In the current search for therapies to improve the symptoms of ASD, it is imperative to have quantitative, objective clinical measures of social deficits in order to evaluate the efficacy of these potential therapies. Eye tracking, a relatively new phenomenon in ASD research, holds promise as a potential measurement tool for evaluation of experimental therapies’ efficacy.  Previous research has suggested that there is a difference of viewing preferences between toddler social scenes and geometric shapes in individuals with ASD compared to neurotypical (NT) peers.

Objectives: Considering that age may impact the saliency of social stimuli, this study aimed to evaluate the viewing preferences of children with ASD and neurotypical peers when given the choice between age-matched social scenes and geometric scenes.

Methods: Eye tracking data was collected for both the ASD and NT participants. Social Communication Questionnaire (SCQ) score was collected for both groups, and Intelligence Quotient (IQ), Social Responsiveness Scale (SRS), Aberrant Behavior Checklist (ABC), and Vineland scores were collected for the ASD children.   The study participants included children ages 6-11 and the eye tracking paradigms presented school-aged children. Each subject viewed three eye tracking paradigms, which were each presented next to a video of moving geometric shapes. The videos were randomly assigned to right versus left to control for a side bias. Total fixation duration of the social scenes and total fixation duration of the geometric scenes were recorded and the ratio of preference was calculated.  The preference ratio was compared between the ASD and NT groups. 

Results:  The final sample consisted of 28 participants with ASD and 18 neurotypical peers.  There were 37% female and 63% male in the study.   Total viewing time of the social scenes significantly differed between the two groups (t= -3.726, p= 0.001), while geometric viewing was not significantly different (t=1.199, p>0.05).  When the ratio between the two types of scenes was compare, a significant difference was found between the ASD and NT groups in their viewing preference (t = -2.534; p< .05). The ASD group demonstrated decreased preference for viewing the social scenes relative to their NT peers.  This relationship was mediated by SCQ score.  As SCQ score increased, there was a decreased preference for viewing the social scenes.  The other testing variables collected did not have a significant impact on the difference between the two groups.

Conclusions: This study found that when social stimuli presented is age-matched, children with ASD still demonstrate a decreased preference for social scenes relative to their NT peers. These findings are similar to the results from a comparable study in which children viewed toddler social scenes versus geometric shapes. Further studies should be conducted to determine if other metrics mediate viewing preferences as SCQ did. Once a better understanding of the social versus geometric scene preference in ASD is established, eye tracking could be a useful tool to evaluate the effectiveness of novel ASD therapies.