A0252
Title: Constrained multivariate functional principal components analysis for eye-tracking experiments
Authors: Brian Kwan - University of California, Los Angeles (UCLA) (United States) [presenting]
Catherine Sugar - UCLA (United States)
Damla Senturk - University of California Los Angeles (United States)
Abstract: Social attention is a neural process relevant to biomarker research in autism spectrum disorder (ASD), and eye-tracking experiments offer extensive insights into attentional processes by providing gaze patterns to sensory stimuli. However, common analysis through summaries, such as total looking time durations in regions of interest, collapses data across trials. A novel multivariate functional outcome is proposed that carries looking time duration information from multiple regions of interest jointly as a function of trial type among static image and dynamic video trials in an Activity Monitoring task without collapsing across trials. A novel constrained multivariate functional principal components analysis is also proposed to capture variation in the proposed outcome, incorporating the constraint in the data that looking time durations from multiple regions of interest need, to sum up to the total trial time. Application to data from the Autism Biomarkers Consortium for Clinical Trials (ABC-CT) study yields new insights into dominant modes of variation of looking time durations from multiple regions of interest for school-age children with ASD and their typically developing peers.
