Modelling Whole-brain Structural Connectivity in Preterm Infants
File(s)
Author(s)
Pandit, Anand
Type
Thesis
Abstract
Premature birth is a major issue for neurological health. Advances in medical care have led to a decrease in neonatal mortality; however, morbidity in preterm survivors remains substantial and neurocognitive issues, particularly involving cognition, language and behaviour are prevalent. Cerebral white matter injury is common in preterm-born infants and is associated with neurocognitive impairments. By identifying the pattern of connectivity changes in the brain following premature birth a more comprehensive understanding of the neurobiology underlying these deficits could be obtained.
Whole-brain macrostructural connectivity was characterised in a group of preterm-born children, and the influence of age and prematurity was explored using a data-driven analysis of diffusion magnetic resonance imaging data. An age-adapted framework, combining anatomical and tissue segmentations with probabilistic diffusion tractography, was used to derive connectivity matrices, weighted by mean tract anisotropy: a measure of connective certainty and strength. In a sparsified, group-consistent connectivity matrix, a novel feature selection method comprising Lasso regression and stability selection was used to identify connections whose mean anisotropy was related to age at imaging or delivery.
As hypothesised, older children were found to have greater connectivity in tracts involving frontal or temporal lobe structures. Increasing prematurity at birth was related to widespread, bilateral reductions in connectivity in all cortical lobes and several sub-cortical structures, consistent with previous histological and neuroimaging data. Results were robust to both increasing sparsity and the removal of subjects with serial imaging in the cohort. The pattern of white matter damage elicited in these results may be responsible for the neurocognitive impairments observed.
This thesis presents a scalable, data-driven method, which could detect contrasting effects of development and prematurity in a sparse model of infant whole-brain structural connectivity. The approach can be used to evaluate connectivity in ever-increasing detail and undertake iterative discovery of the macroconnectome with increasing precision.
Whole-brain macrostructural connectivity was characterised in a group of preterm-born children, and the influence of age and prematurity was explored using a data-driven analysis of diffusion magnetic resonance imaging data. An age-adapted framework, combining anatomical and tissue segmentations with probabilistic diffusion tractography, was used to derive connectivity matrices, weighted by mean tract anisotropy: a measure of connective certainty and strength. In a sparsified, group-consistent connectivity matrix, a novel feature selection method comprising Lasso regression and stability selection was used to identify connections whose mean anisotropy was related to age at imaging or delivery.
As hypothesised, older children were found to have greater connectivity in tracts involving frontal or temporal lobe structures. Increasing prematurity at birth was related to widespread, bilateral reductions in connectivity in all cortical lobes and several sub-cortical structures, consistent with previous histological and neuroimaging data. Results were robust to both increasing sparsity and the removal of subjects with serial imaging in the cohort. The pattern of white matter damage elicited in these results may be responsible for the neurocognitive impairments observed.
This thesis presents a scalable, data-driven method, which could detect contrasting effects of development and prematurity in a sparse model of infant whole-brain structural connectivity. The approach can be used to evaluate connectivity in ever-increasing detail and undertake iterative discovery of the macroconnectome with increasing precision.
Version
Open Access
Date Issued
2013-07
Date Awarded
2013-11
Copyright Statement
Attribution NoDerivatives 4.0 International Licence (CC BY-ND)
Advisor
Edwards, David
Counsell, Serena
Montana, Giovanni
Azzopardi, Denis
Sponsor
Medical Research Council (Great Britain)
Publisher Department
Institute of Clinical Sciences
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)