Systematic differences between ultrasound and MRI assessment of fetal ventricular width in cases of fetal mild ventriculomegaly: a cohort study

Prenatal assessment of the lateral ventricles is essential for detecting and classifying fetal ventriculomegaly (VM). Although ultrasound (US) is the primary screening tool, magnetic resonance imaging (MRI) is increasingly used to complement diagnosis. However, discrepancies between US and MRI measurements, particularly in mild VM, may affect classification, counseling, and planning clinical management. The aim of the study is to compare US and MRI measurements of the fetal lateral ventricles, evaluate the agreement between both methods, with particular focus on mild ventriculomegaly, and additionally assess the feasibility of deriving conversion formulas between both modalities. Methods This cohort study included 120 fetuses examined between 2011 and 2021 at a referral center: 75 with isolated mild VM and 45 controls. All underwent a standardized US assessment and fetal MRI. Ventricular width was measured twice for each modality; mean values were used for statistical comparison. Analysis included Wilcoxon tests, Friedman tests, Bland–Altman plots, Pearson’s chi-square tests, Theil–Sen regression, and linear modeling. Results MRI systematically provided greater ventricular measurements than US. In the study group, the bias was 0.71mm for coronal MRI and 0.33mm for axial MRI. Agreement between methods was limited, with measurement discrepancies observed in 34% of cases (coronal) and 33.3% (axial). Combined analysis of both groups confirmed persistent measurement discrepancies, though the overall axial MRI bias decreased to 0.15mm. Theil–Sen modelling enabled the development of conversion formulas for both measurement planes. Conclusions MRI measurements of the fetal lateral ventricles tend to be slightly higher than ultrasound measurements, particularly in the coronal plane. These systematic differences may change the VM classification and should be considered in clinical interpretation. Ultrasound demonstrated better measurement reproducibility. The obtained formulas may support improved cross-modality comparison in clinical practice.