A machine learning investigation of volumetric and functional MRI abnormalities in adults born preterm

Abstract

Imaging studies have characterized functional and structural brain abnormalities in adults after premature birth, but these investigations have mostly used univariate methods that do not account for hypothesized interdependencies between brain regions or quantify accuracy in identifying individuals. To overcome these limitations, we used multivariate machine learning to identify gray matter volume (GMV) and amplitude of low frequency fluctuations (ALFF) brain patterns that best classify young adults born very preterm/very low birth weight (VP/VLBW; n = 94) from those born full-term (FT; n = 92). We then compared the spatial maps of the structural and functional brain signatures and validated them by assessing associations with clinical birth history and basic cognitive variables. Premature birth could be predicted with a balanced accuracy of 80.7% using GMV and 77.4% using ALFF. GMV predictions were mediated by a pattern of subcortical and middle temporal reductions and volumetric increases of the lateral prefrontal, medial prefrontal, and superior temporal gyrus regions. ALFF predictions were characterized by a pattern including increases in the thalamus, pre- and post-central gyri, and parietal lobes, in addition to decreases in the superior temporal gyri bilaterally. Decision scores from each classification, assessing the degree to which an individual was classified as a VP/VLBW case, were predicted by the number of days in neonatal hospitalization and birth weight. ALFF decision scores also contributed to the prediction of general IQ, which highlighted their potential clinical significance. Combined, the results clarified previous research and suggested that primary subcortical and temporal damage may be accompanied by disrupted neurodevelopment of the cortex.

Keywords: ALFF; VBM; machine learning; multivariate; premature birth; resting-state fMRI.

Figure 1

Voxel probability map of reliable contributions to the FT (full‐term born) versus VP/VLBW (very preterm/ very low birth weight born) group decision boundary using gray matter volume (GMV) of structural MRI data. Voxels with a probability of >50% were overlaid on the single subject MNI template using the MRIcron software package (http://www.mccauslandcenter.sc.edu/mricro/mricron/) [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 2

Voxel probability map of reliable contributions to the FT (full‐term born) versus VP/VLBW (very preterm/ very low birth weight born) group decision boundary using amplitude of low‐frequency fluctuations (ALFF) of resting‐state fMRI data. Voxels with a probability of >50% were overlaid on the single subject MNI template using the MRIcron software package (http://www.mccauslandcenter.sc.edu/mricro/mricron/) [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 3

Voxel probability map of reliable contributions to the FT (full‐term born) versus VP/VLBW (very preterm/ very low birth weight born) group decision boundary using decreased amplitude of low frequency fluctuations (ALFF) of resting‐state fMRI data and VBM. Blue represents decreased VBM and ALFF in both middle temporal gyrus and right insular. Red represents increased VBM and ALFF in right fusiform gyrus and parietal lobes. Yellow represents areas with increased VBM but decreased in ALFF and both superior temporal gyrus and temporal lobes. Violet represents areas with decreased VBM but increased ALFF in subcortex, including the thalamus and caudate nucleus. Voxels with a probability of >50% were overlaid on the single subject MNI template using the MRIcron software package (http://www.mccauslandcenter.sc.edu/mricro/mricron/) [Color figure can be viewed at http://wileyonlinelibrary.com]

Figure 4

Feature selection probability showing the most reliable clinical‐performance contributions to the prediction of VP/VLBW decision scores from the GMV (left) and ALFF (right) brain analyses. Yellow bars indicate positive coefficients (i.e., higher values associated with higher VP/VLBW likeness) and blue bars indicate negative weights (i.e., lower values associated with higher VP/VLBW likeness). Features above a 50% selection threshold most reliably contribute to the models. Abbreviations: BW, birth weight; DNTI, duration of neonatal treatment index; GA, gestational age; INTI, intensity of neonatal treatment index; IQG, full scale intelligence quotient; IQp, performance intelligence quotient; IQv, verbal intelligence quotient; OPTN, optimality score of neonatal treatment; TIV, total intracranial volume [Color figure can be viewed at http://wileyonlinelibrary.com]