Background and Aim Longitudinal studies show that premature birth increases infants’ risk for mental and motor development deficits. Our aim was to investigate the influence of prematurity on cerebral tissue volumes at term obtained with a novel fully automatic segmentation method.
Methods 62 preterm infants (GA 27.7±1.3wks) and 15 term-born infants (GA 40±1.1wks) were scanned at term-equivalent age (GA 40.5±1.5wks). T1 and T2 MR images were segmented with a novel atlas-free automatic method based on morphological constraints. Each brain was separated into the two hemispheres, cortical and subcortical gray matter, myelinated and unmyelinated white matter, brainstem, cerebellum and CSF.
Results Linear regression models were fitted to study the dependency of tissue volumes on GA at birth, GA at scan and intracranial volume. Models show significant dependence on GA at birth for cortical gray matter (Beta=0.270, P=0.000, R2=0.818), unmyelinated white matter (Beta=0.196, P=0.03, R2=0.575), cerebellum (Beta=0.348, P=0.000, R2=0.648) and CSF (Beta= –0.329, P=0.000, R2=0.708).
Wilcoxon Signed Ranks tests showed significantly larger unmyelinated white matter volumes in the right hemisphere compared to the left hemisphere (Z= –4.826, P=0.000), and significantly larger total volumes of the right hemisphere compared to the left hemisphere (Z= –3.486, P=0.000).
Conclusions Reliable volume assessments were derived from the new automatic segmentation. CSF volumes at term increased with lower GA at birth, while cortical gray matter, unmyelinated white matter and cerebellum volumes at term increased with GA at birth, suggesting impaired growth of these tissues associated with prematurity. Cerebral asymmetry was present at term for both preterm and term infants.