Autism spectrum disorder (ASD) is a highly heritable neurodevelopmental disorder affecting more than 1% of children (and almost 2% of boys), with those born in a high risk (HR) family with a child diagnosed with ASD having a ~20 fold greater risk than the general population. Although it can only be reliably diagnosed after the second year of life, ASD originates from a neurodevelopmental mechanism, with symptoms emerging as early as 6 months of age. Prospective large-scale longitudinal neuroimaging studies like the Infant Brain Imaging Study (IBIS), of those born into a high-risk (HR) family, are extremely valuable as they facilitate discovery of the earliest manifestations of ASD. The overarching goal of this proposal is a comprehensive comparative longitudinal analysis of brain organization of high- and low-risk populations, by using IBIS diffusion MRI (dMRI) data collected at 3 time points in the first 2 years of life, to elucidate the earliest manifestations of ASD and disorder trajectory. In addition, we aim to create a biomarker of ASD risk from brain connectivity features and their developmental trajectories, and to understand the brain bases of familial risk (independent of ASD). dMRI provides an insight into the structural organization of the brain represented as a connectome. ?Miswiring? of the structural connectome can manifest as neuro-immaturity of WM and changes in network structures, that is, subnetworks (collection of strongly inter-connected regions associated with a distinct communication pattern), and the communication backbone (overall architecture of communication between subnetworks). Both connectivity and WM quality can be compromised in ASD compared to controls, suggesting a ?miswired? connectome. Studying ASD related early developmental changes in connectivity requires design of novel analysis methods based on ?longitudinal? connectomic features that are 4D features that incorporate their temporal evolution over development, culminating in the creation of novel imaging-based biomarkers.
In Aim 1, we will create a new method of extracting longitudinal fiber tracts to identify and analyze differences between the developmental tract trajectories in LR-, HR- and HR+ subjects using geometry and quality features derived from the fiber bundles, to identify differences in neuro-immaturity.
In aim 2, we will design novel methods to extract longitudinal network structures like structurally-cohesive and functionally-defined subnetworks and the global communication backbone, and investigate developmental differences modulated by familial risk and gender. Finally, in Aim 3 we will use the tract and network features extracted from Aims 1 and 2 to develop imaging based markers that will assist with early ASD prediction, identifying siblings who could gain from an early therapeutic intervention and provide an index of development in the form of brain connectivity ?maturational age? to help characterize developmental delays and brain patterns corresponding to the same. The longitudinal analysis methods developed in the proposal will be generalizable to other clinical populations that can gain from longitudinal analysis spanning several time points.

Public Health Relevance

The project aims at comprehensive comparative analysis of diffusion-based structural connectivity in high- and low-risk populations, to elucidate the earliest manifestation of Autism Spectrum Disorder (ASD). This will require the design of novel sophisticated longitudinal analysis methods based on longitudinal tracts and network structures that inherently incorporate the temporal evolution over development. Finally, using the developmental trajectories of these features, we aim to create imaging-based biomarkers of familial risk, development and predictive of ASD prior to the age of diagnosis.

National Institute of Health (NIH)
Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD)
Research Project (R01)
Project #
Application #
Study Section
Child Psychopathology and Developmental Disabilities Study Section (CPDD)
Program Officer
Kau, Alice S
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Pennsylvania
Schools of Medicine
United States
Zip Code
Fortin, Jean-Philippe; Parker, Drew; Tunç, Birkan et al. (2017) Harmonization of multi-site diffusion tensor imaging data. Neuroimage 161:149-170