EpiBio Seminar

Wright Hall 150

12-1 p.m.

Improving estimates of brain connectivity in autism spectrum disorder

Ben Risk, PhD
Associate Professor, Department of Biostatistics and Bioinformatics
Emory University

Join us for a presentation from Dr. Ben Risk of Emory University.

The seminar’s abstract is as follows: Autism spectrum disorder is a neurodevelopmental condition characterized by difficulties with social interactions, communication, and restricted or repetitive behaviors. Resting-state functional magnetic resonance imaging is used to examine the interactions between brain regions, known as functional connectivity. Head motion during a scan creates artifacts that distort functional connectivity. Early studies reported altered functional connectivity in autistic children. However, these differences may have been biased because autistic children tend to move more. More recent studies in pediatric neuroimaging remove high motion participants. However, children with more severe symptoms tend to move more, so excluding high-motion participants creates selection bias. In this talk, I will discuss how my lab is designing experiments and developing statistical methods to better characterize functional connectivity in autistic children. First, we flexibly model the impacts of motion using an ensemble of machine learning methods. We propose a causal mediation framework, where we treat motion as a stochastic intervention. Second, I will present sparse independent component analysis, which can be applied to the estimation of resting-state networks to reduce noise and improve interpretability. Finally, I will discuss our ongoing study designed to include autistic children across the full autism spectrum, including those with more severe symptoms who are typically excluded from connectivity studies.