Microstructure of Saccadic Behavior in High and Low-Risk Infants: Visual Scanning of Static Novel Faces

Background: Accessing and efficiently processing information from faces remains a compelling assay of social cognition in infancy.  Previous research suggests that children and adults with autism use atypical face scanning strategies to acquire social information necessary to navigate complex social situations.  Additionally, some research suggests that atypical face scanning strategies may be present in infants prior to a diagnosis and may represent an indicator of risk for an eventual diagnosis.

Objectives: As part of the IBIS Network investigating early brain development in infant siblings of children with autism at 6, 12, and 24 months, we designed an eye-tracking battery to assess basic cognitive and social-cognitive skills that may develop on an atypical trajectory in children that eventually meet diagnostic criteria for autism.  One aspect of this battery is examining the saccade dynamics elicited when high- and low-risk infants visually scan unfamiliar faces.  In addition to total time spent in specific areas of interest (AOI) such as the eye region and mouth region, we are particularly interested in the number of discreet fixations and the average duration of fixation in specified AOIs.

Methods: Eye-tracking data were collected via a Tobii 1750 eye-tracker.  The face scanning task involves the presentation of static faces showing positive emotions for a short duration of time (4 seconds) in order to maximize scanning behavior that represents an infant’s strategy for acquiring salient aspects of a face encountered for the first time.  In addition to measuring total gaze time, we extracted the number of discreet fixations and the average duration of fixation on specific AOIs.  Associations between general cognitive functioning via the Mullen Early Learning Scales and markers of autism via the Autism Observation Scale for Infants (AOSI) will also be reported.

Results: Our current sample size [n = 12 high-risk sibs (9 six month-olds and 3 twelve month-olds), n = 7 low-risk sibs (6 six month-olds and 1 twelve month-old)] will nearly double over the next couple of months and we plan to present data from ~40 infants at IMFAR.  Preliminary analyses indicate that that total scanning time of the face is equivalent for both groups (p = 0.925).  There are also substantial individual differences in performance on the face scanning task in both groups, which will inform associations with the Mullen and AOSI as the n continues to increase.  We will also accrue a number of follow-up assessments and data on ~20 children will be presented who have data at both 6 and 12 months of age.
Conclusions: As we follow these children longitudinally we expect to see patterns of specific deviations from the normal distribution that may lead to specific deficits as the children enter the second and third years of life.  Additionally, we will be able to determine whether aspects of saccadic behavior and visual attention might serve as a compensatory mechanism that facilitates typical information processing strategies.