Childhood Outcomes of Preterm Brain Abnormalities

Overview

Prematurely born children are at higher risk of cognitive impairments and behavioral disorders than full-term children. There is growing evidence of significant volumetric and shape abnormalities in subcortical structures of premature neonates, which may be associated to negative long-term neurodevelopmental outcomes. The general objective is to look directly at the long-term neurodevelopmental implications of these neonatal subcortical structures abnormalities. Investigators propose to develop biomarkers of prematurity by comparing the morphological and diffusion properties of subcortical structures between preterm, with and without associated brain injuries, and full-term neonates using brain MRI. By combining subcortical morphological and diffusion properties, investigators hypothesize to be able to: (1) delineate specific correlative relationships between structures regionally and differentially affected by normal maturation and different patterns of white matter injury, and (2) improve the specificity of neuroimaging to predict neurodevelopmental outcomes earlier. The specific aims and general methodology are: 1) Build a new toolbox for neonatal subcortical structures analyses that combine a group lasso-based analysis of significant regions of shape changes, a structural correlation network analysis, a neonatal tractography, and tensor-based analysis on tracts; 2) Ascertain biomarkers of prematurity in neonates with different patterns of abnormalities using correlational and connectivity analysis within and between structures features; 3) Assess the predictive potential of subcortical imaging on neurodevelopmental outcomes by correlating neonatal imaging results with long-term neurodevelopmental scores at 9 and 18 months, and 6-8 years, follow-up. In each of these aims, investigators will use advanced neuroimaging analysis developed by their group and collaborator, including multivariate tensor-based morphometry and multivariate tract-based analysis. This application will provide the first complete subcortical network analysis in both term and preterm neonates. In the first study of its kind for prematurity, investigators will use sparse and multi-task learning to determine which of the biomarkers of prematurity at birth are the best predictors of long-term outcome. Once implemented, these methods will be available to compare subcortical structures for other pathologies in newborns and children.

The last months of pregnancy are particularly important for the development of the child's brain, and the consequences of premature birth on its development can be substantial. Prematurely born children are at higher risk of various cognitive impairments and exhibits more behavioral disorders than full-term born children. Thus early detection and management of at risk children are essential. There is growing evidence of significant volumetric abnormalities in subcortical structures of premature neonates, which may be associated to negative long-term neurodevelopmental outcomes. Understanding these abnormalities could help elucidate the underlying pathophysiology and enable early determination of at-risk patients, both of which would inform the design of novel treatment strategies. However, to date there is still a lack of sensitive, reliable, and accessible algorithms capable of characterizing the influence of prematurity on the anatomy of neonatal brain subcortical structures. In addition, few studies have looked directly at the long-term neurodevelopmental implications of these neonatal subcortical structures abnormalities. Predicting long-term neurodevelopmental outcomes early on - and preferably at neonatal ages - is likely to have a transformative effect on their outcome. Our preliminary data indicate significant morphological differences in the putamen, ventricles, corpus callosum, and thalamus between preterm and term neonates. Investigators propose to develop biomarkers of prematurity by statistically comparing the morphological and diffusion properties of subcortical structures between preterm and term neonates using brain MRI. These results will further be used in a sparse learning framework to predict long-term neurodevelopmental outcomes of prematurity. Hypotheses: By combining subcortical morphological and diffusion properties, we will be able to: (1) delineate specific correlative relationships between structures regionally and differentially affected by normal maturation and different patterns of white matter injury, and (2) improve the specificity of neuroimaging to predict neurodevelopmental outcomes earlier. Aim 1: Build a new toolbox for neonatal subcortical structures analyses that combine 1) a group lasso-based analysis of significant regions of shape changes, 2) a structural correlation network analysis, 3) a neonatal tractography, and 4) tensor-based analysis on tracts. Aim 2: Ascertain biomarkers of prematurity in neonates with different patterns of abnormalities. Aim 3: Assess the predictive potential of imaging and clinical features on neurodevelopmental outcomes among premature children at 9 and 18 months and 6-8 years of age. Impact: This application will provide the first complete subcortical network analysis in both term and preterm neonates. In the first study of its kind for prematurity, investigators will use sparse and multi-task learning to determine which of the biomarkers of prematurity at birth are the best predictors of long-term outcome. The expected findings could improve the ability to predict these outcomes and enable the design of early treatments - before years of pathological brain development and symptoms occur.