Application of grey matter-based spatial statistical analysis methods in neonatal cerebral cortical development
Objective:To explore developmental changes in microstructural properties of the cerebral cortex during the neonatal period using gray matter-based spatial statistics (GBSS). Materials and Methods:From January 2011 to June 2013,73 neonates who underwent MRI examinations from our hospital were retrospectively recruited. The cortical skeleton was extracted and the diffusion tensor imaging (DTI) covariates fractional anisotropy (FA),and mean diffusivity (MD) were projected onto the cortical skeleton using grey matter-based spatial statistics. Bilaterally symmetrical cortical regions of interest (ROIs) were selected based on neonatal T1WI atlas of brain regions and the mean values of DTI covariates in the frontal,temporal,parietal and occipital cortex as well as the mean values of the corresponding covariates within each ROI were calculated,and the values of DTI covariates in the frontal,temporal,parietal and occipital cortex as well as within each ROI were further correlated with gestational age,corrected birth weight,crown-heel length and head circumference. The correlations between gestational age and birth indicators and cortical FA and MD parameters were statistically analyzed using a general linear model. Results:Cortical ROI-based analysis revealed that changes in neonatal cortical FA were based on a turning point of 38 weeks of gestational age. Before the turning point,there was no significant correlation between neonatal frontal,temporal,parietal,and occipital cortical FA and gestational age,and after the turning point,only the parietal cortical FA (r=0.424,P=0.009) increased with increasing gestational age. Based on all 48 cortical ROIs and voxel level analyses,FA values of right superior temporal gyrus,insular cortex,middle frontal gyrus,cingular gyrus,bilateral lateral fronto-orbital gyrus,inferior frontal gyrus,postcentral gyrus,precentral gyrus,and left superior parietal lobule were found to be positively correlated with gestational age,whereas MD values were negatively correlated with gestational age (P<0.05). The highest correlation between FA values and gestational age was found in the right postcentral gyrus (PoCG) (r=0.628,P=0.032),and the highest correlation between MD values and gestational age was found in the left PoCG (r=−0.598,P=0.041). Both voxel and ROI based analyses showed that no significant correlation was found between birthweight,head circumference,crown-heel length and neonatal cortical FA and MD values. Conclusions:The neonatal cerebral cortex is characterized by regional heterogeneity,with primary sensory and motor cortex maturing earlier than the association cortex. Compared with birth weight,crown-heel length and head circumference,gestational age is the main index affecting the development of neonatal cerebral cortex.
万方数据
维普
