Biology subjects

Adamson, C. L.

Publications and source records attributed to Adamson, C. L..

2 recordsLinked to original sources

bioRxiv2024

Heterogeneous, temporally consistent, and plastic brain development after preterm birth

The current view of neurodevelopment after preterm birth presents a strong paradox: diverse neurocognitive outcomes suggest heterogeneous neurodevelopment, yet numerous brain imaging studies focusing on average dysmaturation imply largely uniform aberrations across individuals1. Here we show both, spatially heterogeneous individual brain abnormality patterns (IBAPs) but with consistent underlying biological mechanisms of injury and plasticity. Using cross-sectional structural magnetic resonance imaging data from preterm neonates and longitudinal data from preterm children and adults in a normative reference framework2, we demonstrate that brain development after preterm birth is highly heterogeneous in both severity and patterns of deviations. Individual brain abnormalities were also consistent for their extent and location along the life course, associated with glial cell underpinnings, and plastic for influences of the early social environment. Thus, IBAPs of preterm birth are spatially heterogenous, temporally consistent for extent, spatial location, and cellular underpinnings, and plastic for social-environmental impacts. Our findings extend conventional views of preterm neurodevelopment, revealing a nuanced landscape of individual variation, with consistent commonalities between subjects. This integrated perspective of preterm neurodevelopment implies more targeted theranostic intervention strategies, specifically integrating brain charts2 and imaging at birth, as well as social interventions during early development3.

neuroscience↗

bioRxiv2021

Benchmarking Geometric Deep Learning for Cortical Segmentation and Neurodevelopmental Phenotype Prediction

The emerging field of geometric deep learning extends the application of convolutional neural networks to irregular domains such as graphs, meshes and surfaces. Several recent studies have explored the potential for using these techniques to analyse and segment the cortical surface. However, there has been no comprehensive comparison of these approaches to one another, nor to existing Euclidean methods, to date. This paper benchmarks a collection of geometric and traditional deep learning models on phenotype prediction and segmentation of sphericalised neonatal cortical surface data, from the publicly available Developing Human Connectome Project (dHCP). Tasks include prediction of postmenstrual age at scan, gestational age at birth and segmentation of the cortical surface into anatomical regions defined by the M-CRIB-S atlas. Performance was assessed not only in terms of model precision, but also in terms of network dependence on image registration, and model interpretation via occlusion. Networks were trained both on sphericalised and anatomical cortical meshes. Findings suggest that the utility of geometric deep learning over traditional deep learning is highly task-specific, which has implications for the design of future deep learning models on the cortical surface. The code, and instructions for data access, are available from https://github.com/Abdulah-Fawaz/Benchmarking-Surface-DL.

neuroscience↗