Cortical microstructures at birth predict neurodevelopmental outcome at 2 years of age

Cortical microstructural architecture in preterm brain, quantitatively delineated with DTI fractional anisotropy (FA), is characterized by various levels of dendritic arborization. With cortical FA an important indicator of cortical maturational level, we hypothesized that cortical FA at birth predicts neurodevelopmental outcome at 2 years of age. We used FA across cortical regions of neonate brains to predict their Bayley scores at 2 years of age with machine learning algorithms. A highly predictive model was achieved with inhomogeneous pattern across cortical areas. Significant correlation was found between FA values in widespread cortical regions at birth and cognitive scores at 2 years.