Non-negative data-driven mapping of structural connections with application to the neonatal brain

E Thompson,A.R Mohammadi-Nejad,E.C Robinson,J.L.R Andersson,S Jbabdi,M.F Glasser,M Bastiani,S.N Sotiropoulos

doi:10.1016/j.neuroimage.2020.117273

Abstract

Mapping connections in the neonatal brain can provide insight into the crucial early stages of neurodevelopment that shape brain organisation and lay the foundations for cognition and behaviour. Diffusion MRI and tractography provide unique opportunities for such explorations, through estimation of white matter bundles and brain connectivity. Atlas-based tractography protocols, i.e. a priori defined sets of masks and logical operations in a template space, have been commonly used in the adult brain to drive such explorations. However, rapid growth and maturation of the brain during early development make it challenging to ensure correspondence and validity of such atlas-based tractography approaches in the developing brain. An alternative can be provided by data-driven methods, which do not depend on predefined regions of interest. Here, we develop a novel data-driven framework to extract white matter bundles and their associated grey matter networks from neonatal tractography data, based on non-negative matrix factorisation that is inherently suited to the non-negative nature of structural connectivity data. We also develop a non-negative dual regression framework to map group-level components to individual subjects. Using in-silico simulations, we evaluate the accuracy of our approach in extracting connectivity components and compare with an alternative data-driven method, independent component analysis. We apply non-negative matrix factorisation to whole-brain connectivity obtained from publicly available datasets from the Developing Human Connectome Project, yielding grey matter components and their corresponding white matter bundles. We assess the validity and interpretability of these components against traditional tractography results and grey matter networks obtained from resting-state fMRI in the same subjects. We subsequently use them to generate a parcellation of the neonatal cortex using data from 323 new-born babies and we assess the robustness and reproducibility of this connectivity-driven parcellation.

Highlights

The neonatal period is a critical time for brain development, during which the refinement and maturation of white matter connections lay the groundwork for later cognitive development (Ball et al, 2015; Counsell et al, 2008; Girault et al, 2019)
DMRI tractography protocols for identifying specific white matter bundles typically rely on delineation of regions of interest (ROIs) that provide a priori anatomical knowledge on the route of the tract; these ROIs can be defined relative to a template for automated delineation (Bastiani et al, 2019; De Groot et al, 2013; Warrington et al, 2020)
We explore the validity and interpretability of i) the automatically detected white matter patterns against results from standard tractography protocols available through the developing Human Connectome Project (dHCP) (Bastiani et al, 2019) and ii) the grey matter patterns against components obtained from data-driven mapping of resting-state functional MRI (fMRI) in the same subjects

Summary

Introduction

The neonatal period is a critical time for brain development, during which the refinement and maturation of white matter connections lay the groundwork for later cognitive development (Ball et al, 2015; Counsell et al, 2008; Girault et al, 2019). With diffusion MRI (dMRI) we can track these connections non-invasively and in vivo, which enables us to study the early development of structural connectivity and microstructure, even during the first weeks of life (see (Ouyang et al, 2019) for a recent review). DMRI studies of neonates have shown that the trajectory of fibre maturation reflects the neurodevelopmental hierarchy, with primary motor and sensory tracts developing earlier than the association tracts that enable higher order functioning (Dubois et al, 2008; Kulikova et al, 2015; Partridge et al, 2004). DMRI tractography protocols for identifying specific white matter bundles typically rely on delineation of regions of interest (ROIs) that provide a priori anatomical knowledge on the route of the tract; these ROIs can be defined relative to a template for automated delineation (Bastiani et al, 2019; De Groot et al, 2013; Warrington et al, 2020)

Methods

Results

Conclusion