18150
Impaired Maturational Changes of Network Organization in ASD: An Ica Study Using Resting State fMRI

Friday, May 16, 2014
Atrium Ballroom (Marriott Marquis Atlanta)
M. Sullivan, I. Fishman, Y. Cabrera and R. A. Müller, Brain Development Imaging Laboratory, Dept. of Psychology, San Diego State University, San Diego, CA
Background:   Converging evidence indicates that brain abnormalities in autism spectrum disorders (ASD) involve atypical network connectivity, yet the particular connectivity patterns and the extent to which they deviate from those seen in typical development are currently debated. While there is evidence supporting both greater and weaker brain network connectivity in individuals with ASD, as compared to their typically developing (TD) peers, the developmental framework emphasizing age-associated changes in connectivity is often overlooked, or altogether ignored. 

Objectives:   To examine developmental trajectories of network connectivity in a sample of children and adolescents with ASD, as compared to TD controls.  Specifically, recent findings suggest that, while typical development is associated with increasingly segregated (weaker between-network connectivity) and increasing integrated (greater within-network connectivity) brain networks, this pattern appears to be disrupted in ASD.  The aim of this project was examine this notion using a data-driven, model-free approach to quantify large-scale patterns of functional connectivity (i.e., functional networks) in children and adolescents with ASD. 

Methods:   Resting-state functional MRI (rs-fMRI) data (acquired for 6:10 minutes) from 28 children and adolescents with ASD (ages 7-18 years old) and 33 TD participants, matched on age, gender, non-verbal IQ, and head motion, were included.  Utilizing publicly available templates of major functional brain networks (Smith et al., 2009) generated through Independent Component Analysis (ICA) of rs-fMRI data, we applied the dual regression approach implemented in FSL to decompose each participant’s BOLD data into 10 functional networks, also known as independent components.  The dual regression algorithm was applied to identify both participant-specific time courses and spatial maps corresponding to the 10 brain networks.  Z scores associated with each network were then compared between groups (with a two-sample t-test), as well as correlated with age within each group.

Results:   We observed significantly different functional connectivity within several large-scale brain networks in children and adolescents with ASD, relative to TD controls.  Namely, greater connectivity was found for the visual/medial, visual/occipital pole and the cerebellar networks.  Moreover, we found differential age slopes of functional connectivity between TD and ASD groups. Some networks (e.g., two visual networks) showed significant reduction in connectivity with age in the TD group, but no association with age in ASD; others (such as default mode network) showed significantly increasing connectivity with age in TD participants, but no age effect in ASD.

Conclusions:   Our findings suggest that large-scale functional brain networks have differential maturational trajectories in TD children and adolescents, with decreasing integration in visual networks, but increasing integration in DMN. The trajectories of network maturation appear to be disturbed pervasively in ASD, with many networks showing no significant age-related changes. Our findings underscore the importance of considering developmental trends and trajectories when characterizing brain network abnormalities in ASD.