Using a Portable Eye-Tracking Device to Measure Social Impairment in Children with Autism Spectrum Disorder

Background: Autism spectrum disorder (ASD) is a neurodevelopmental condition affecting 1 in 68 children and carrying a public health burden of $69 billion per year. Available ASD treatments are reliant on early and accurate identification. Expert clinicians can reliably diagnosed ASD by age two; nonetheless, research has found that a diagnosis of ASD for underserved and minority populations can be delayed several years despite parental concerns. Using eye-tracking measures, visual attention in ASD has been the focus of extensive research as an early and quantifiable marker of the disease. Indeed, diminished visual social attention and atypical non-social engagement have been correlated with increased ASD symptomatology.

Objectives: The overall objective of this study is to evaluate the utility of a tablet-based eye tracking paradigm to asses social and non-social engagement in children with ASD as a diagnostic aid in community settings. We will also examine the relationship between gaze preference to social and non-social scenes with measures of social impairment, cognitive and adaptive functioning.

Methods: Twelve children with ASD, and 11 age-matched typically developing (TD) controls were included in this study. Children with ASD were diagnosed based on psychiatric evaluation, DSM-5 criteria, and the ADOS-2. All participants received measures of cognitive function. In addition, children with ASD were administered the Vineland Adaptive Behavior Scale- 2^nd edition (VABS-2). All participants completed the eye-tracking paradigm, which consists of a 50 second video in which a social scene and a non-social scene are displayed side-by-side. The algorithm measures time spent looking at: (1) social scene, (2) non-social scene, (3) distraction (i.e. middle of the screen, outside screen frame). A subset of TD and ASD participants (n=11) completed the Social Responsiveness Scale 2^nd Edition (SRS-2).

Results: Gaze preference for the social scene was significantly higher for the TD group as compared to ASD participants (ASD: M=16.58, SD=12.2; TD: M=33.55, SD=16.36; p=0.01). There was no difference in time spent looking at the non-social scene (ASD: M=18.63, SD=13.01; TD: M=13.01, SD=11.27; p=0.317) or the distraction condition (ASD: M=64.75 SD=19.72; TD: M=52.88, SD=21.94; p=0.223). Gaze preference towards the social scene was not correlated with the adaptive behavior composite on the VABS-2 (r = 0.108, p = 0.505) or with full IQ scores (r = 0.293, p=0.197). SRS-2 total scores and time spent looking at the social scene had a negative relationship approaching significance (r=-0.592, p = 0.055).

Conclusions: Consistent with the existing literature, we found that children with ASD spent significantly less time looking at the social scene as compared to their TD peers. Importantly, gaze preference was not impacted by measures of cognitive or adaptive functioning. Interestingly, our results suggest that social impairment as measured by the SRS-2 is inversely correlated with gaze preference towards the social scene, suggesting that the eye-tracking paradigm might be sensitive social functioning in children with ASD and, as such, useful as a measure of treatment response. Overall, these results highlight the utility of the application and warrant continued larger-scale investigation.