Autism spectrum disorder (ASD) is defined behaviorally, by persistent difficulties with social communication, plus restricted, repetitive behaviors; its neural bases are not fully understood. ASD is also highly heterogeneous, in some cases a mild condition, but in others severely disabling. Currently there are no biomarkers used to identify young children with ASD, or to stratify them as more or less severe. Rather, reliable diagnosis at a young age requires observation by extensively trained clinician specialists. This dissertation is in part an attempt to contribute to the search for neural and behavioral valid markers of ASD using eye tracking and EEG.

Studies 1 and 2 examine test performance on eye tracking tasks that pair social images and geometric images as an objective indicator of ASD in 12-48 months infants/toddlers. Study 1 examines sex differences in social attention and whether eye tracking during presentation of competing geometric and social images is equally effective for detecting ASD in female and male toddlers. Increased social attention could serve to protect females from reaching diagnostic thresholds for ASD, impacting female rates of diagnosis. The hypothesis that high risk females (non-ASD but with a diagnosed older sibling) exhibit increased social attention is also explored. Study 2 examines the stability of eye tracking test performance as an ASD indicator when images of biological motion (kids dancing) are replaced with more complex social interactions and a range of positive and negative emotional expressions. Symptom severity differences between ASD subtypes defined by eye tracking results are shown. Combining multiple such eye tracking tests is proposed to improve sensitivity while maintaining high specificity. Study 3 considers EEG mu suppression reflecting mirroring (recruiting first person sensorimotor representations for simulating others during social perception) as a potential neural biomarker for ASD. This study of neurotypical adults was the first to show mirroring of emotional faces reflected in mu suppression. How this finding might lead to a mu suppression neural biomarker for detecting ASD at the individual subject level is discussed also. Collectively, these studies explore potential autism biomarkers that might one day be used to efficiently and objectively identify and stratify toddlers with ASD.