B0307
Title: Hybrid principal components analysis for region-referenced longitudinal functional EEG data
Authors: Damla Senturk - University of California Los Angeles (United States) [presenting]
Aaron Scheffler - UCLA (United States)
Donatello Telesca - UCLA (United States)
Qian Li - UCLA (United States)
Catherine Sugar - UCLA (United States)
Charlotte DiStefano - UCLA (United States)
Shafali Jeste - UCLA (United States)
Abstract: The electroencephalography (EEG) data produce data frames of complex structure that includes functional, longitudinal, and regional dimensions. Our motivating example is a word segmentation paradigm where typically developing (TD) children and children with Autism Spectrum Disorder (ASD) were exposed to a continuous speech stream. For each subject, continuous EEG signals recorded were transformed into the frequency domain resulting in region-referenced principal power where one-second segments throughout the experiment represent the longitudinal dimension, principal power obtained across frequencies represent the functional dimension and the scalp regions represent the regional dimension. We propose a hybrid principal components analysis (HPCA) for region-referenced longitudinal functional EEG data which utilizes both vector and functional principal components analyses and does not collapse information along any of the three dimensions of the data. The proposed decomposition only assumes weak separability of the higher-dimensional covariance process and utilizes a product of one dimensional eigenvectors and eigenfunctions, obtained from the frequency, segment, and region marginal covariances, to represent the observed data, providing a computationally feasible nonparametric decomposition. A mixed effects modeling framework is proposed to estimate the model components, coupled with a bootstrap test for group level inference; both geared towards sparse data applications.
