Quantitative assessment of brain sparing effect in fetal growth restriction: structural evidence from fetal MRI

Abstract

Objective To differentiate brain-sparing patterns in fetal growth restriction (FGR) versus small-for-gestational-age (SGA) fetuses using MRI volumetry. Methods This retrospective study analyzed whole-body MRI data from 378 fetuses (28–40 weeks’ gestation) collected between 2018 and 2025, comprising 308 normal controls, 31 SGA, and 39 FGR cases. Fetal body and brain volumes were segmented using a nnU-Net model. Adjusted linear regression models analyzed growth trajectories of body volume, brain volume, and the brain-to-body volume ratio. Diagnostic performance was evaluated using the area under the curve (AUC). Results SGA fetuses exhibited an 18.2% lower body growth rate than controls (156.1 vs. 190.9 mL/week) but comparable brain growth, indicating preserved brain-sparing. Conversely, FGR fetuses showed significant reductions in both brain (−18.4%) and body (−15.8%) growth rates. Consequently, the brain-to-body volume ratio decreased rapidly in FGR (−0.005/week vs. −0.002/week in controls), signaling a failure of the brain-sparing mechanism. The ratio model achieved AUCs of 0.989 (Normal vs. FGR), 0.94 (Normal vs. SGA), and 0.883 (SGA vs. FGR). Conclusion SGA is characterized by restricted body growth with relative brain preservation, confirming active brain-sparing. FGR involves significant global restriction and a disproportionately decreasing volume ratio, signifying a breakdown of this mechanism. The brain-to-body volume ratio is a sensitive biomarker for distinguishing these distinct adaptive patterns.