Gaze Modification Strategies for Toddlers with ASD

Yale University, New Haven CT

Investigators

Linked publications & trials

Abstract

Abstract Recent eye-tracking studies have shown that 20-month old toddlers with autism spectrum disorders (ASD), as compared to typically developing (TD) and developmentally-delayed (DD) toddlers, attend less to the shared activities of others1 as well as to faces trying to engage them using direct eye-contact and child-directed speech2 (i.e. bids for dyadic engagement). Since attending to others, and especially their actions, is critically linked to both social skill development and observational learning, teaching toddlers and children with ASD normative strategies for viewing others and their activities may provide new opportunities for them to learn about their social world. This proposal initiates a highly novel line of research which uses adaptive eye-tracking technology and non-adaptive perceptually augmented cues in order to help 20-month-old toddlers allocate their attention to people and their actions in a more typical fashion. The project will be conducted in two phases. In the first phase, normative gaze patterns will be collected from 18-24 month old typically developing (TD) toddlers (N=50) watching videos of actresses engaged in everyday activities, with an emphasis on scenarios incorporating bids for dyadic engagement and shared activities. In the second phase, 20-month-old toddlers with a preliminary diagnosis of ASD will be shown videos in the following 3 conditions: (1) a non-adaptive cue condition (N=16), where perceptual enhancement will be employed at selected scene locations (e.g. increasing contrast or emphasizing motion) in order to maintain the viewers' attention towards normative locations of scanning~ (2) an adaptive cue condition (N=16), where, as viewers deviate from the normative gaze pattern, the scene automatically redirects attention back towards the normative pattern by darkening or deemphasizing atypical locations on the scene~ and (3) a control condition (N=16), where toddlers will view the scenes without manipulation, as is shown to TD participants in the first phase. Manipulated blocks will be preceded and followed by dyadic bid scenes and activity scenes which will establish both baseline scanning patterns and the effects of gaze manipulation. Behavioral assessments will be employed before and after the eye-tracking sessions in order to gauge the generalizability of any changes to visual scanning strategies. If successful, this study may point towards new methods for training attention towards social information in toddlers with ASD, and may also lead to new eye-tracking paradigms that may be more sensitive to atypical patterns of attention.