Machine-learning to characterise neonatal functional connectivity in the preterm brain.

Brain development is adversely affected by preterm birth. Magnetic resonance image analysis has revealed a complex fusion of structural alterations across all tissue compartments that are apparent by term-equivalent age, persistent into adolescence and adulthood, and associated with wide-ranging neurodevelopment disorders. Although functional MRI has revealed the relatively advanced organisational state of the neonatal brain, the full extent and nature of functional disruptions following preterm birth remain unclear. In this study, we apply machine-learning methods to compare whole-brain functional connectivity in preterm infants at term-equivalent age and healthy term-born neonates in order to test the hypothesis that preterm birth results in specific alterations to functional connectivity by term-equivalent age. Functional connectivity networks were estimated in 105 preterm infants and 26 term controls using group-independent component analysis and a graphical lasso model. A random forest-based feature selection method was used to identify discriminative edges within each network and a nonlinear support vector machine was used to classify subjects based on functional connectivity alone. We achieved 80% cross-validated classification accuracy informed by a small set of discriminative edges. These edges connected a number of functional nodes in subcortical and cortical grey matter, and most were stronger in term neonates compared to those born preterm. Half of the discriminative edges connected one or more nodes within the basal ganglia. These results demonstrate that functional connectivity in the preterm brain is significantly altered by term-equivalent age, confirming previous reports of altered connectivity between subcortical structures and higher-level association cortex following preterm birth.

MRI Findings at Term-Corrected Age and Neurodevelopmental Outcomes in a Large Cohort of Very Preterm Infants.

BACKGROUND AND PURPOSE: Brain MR imaging at term-equivalent age is a useful tool to define brain injury in preterm infants. We report pragmatic clinical radiological assessment of images from a large unselected cohort of preterm infants imaged at term and document the spectrum and frequency of acquired brain lesions and their relation to outcomes at 20 months. MATERIALS AND METHODS: Infants born at <33 weeks' gestation were recruited from South and North West London neonatal units and imaged in a single center at 3T at term-equivalent age. At 20 months' corrected age, they were invited for neurodevelopmental assessment. The frequency of acquired brain lesions and the sensitivity, specificity, and negative and positive predictive values for motor, cognitive, and language outcomes were calculated, and corpus callosal thinning and ventricular dilation were qualitatively assessed. RESULTS: Five hundred four infants underwent 3T MR imaging at term-equivalent age; 477 attended for assessment. Seventy-six percent of infants had acquired lesions, which included periventricular leukomalacia, hemorrhagic parenchymal infarction, germinal matrix-intraventricular hemorrhage, punctate white matter lesions, cerebellar hemorrhage, and subependymal cysts. All infants with periventricular leukomalacia, and 60% of those with hemorrhagic parenchymal infarction had abnormal motor outcomes. Routine 3T MR imaging of the brain at term-equivalent age in an unselected preterm population that demonstrates no focal lesion is 45% sensitive and 61% specific for normal neurodevelopment at 20 months and 17% sensitive and 94% specific for a normal motor outcome. CONCLUSIONS: Acquired brain lesions are common in preterm infants routinely imaged at term-equivalent age, but not all predict an adverse neurodevelopmental outcome.