Normative range of MRI-based fetal body volume and association with ultrasonographically estimated fetal weight at 16-36 weeks: prospective study.

Abstract

The aims of this study were to define normative values of fetal body volume (FBV) measured using magnetic resonance imaging (MRI) from 16 to 36 gestational weeks (GW), to assess our MRI-based estimated fetal weight (EFW) using ultrasound (US)-based EFW as reference, and to compare it with EFW derived from previously published MRI- or computed tomography (CT)-based formulae. This single-center prospective substudy of the ongoing Lumiere on the Fetus trial included healthy pregnant women with a healthy, singleton, 16-36-week fetus evaluated between December 2021 and June 2022. FBV was measured using MRI with manual segmentation of T2-weighted images. Based on our previous work, we defined MRI-FBV as being equal to MRI-derived EFW. These measurements were compared to US-EFW that was determined using Hadlock's formula < 7 days prior to MRI, by calculating the difference between them (MRI-EFW - US-EFW) and the relative error of MRI-EFW using US-EFW as the gold standard ((MRI-EFW - US-EFW)/US-EFW) × 100)). We also compared with US-EFW the performance of a further four previously published MRI- or CT-based formulae to estimate fetal weight. The study cohort comprised 260 healthy pregnant women with a healthy singleton fetus. FBV was measured using MRI in all 260 cases and we present these normative values. US-EFW was obtained < 7 days prior to MRI in 196 (75.38%) cases. All except four (1.54%) fetuses were followed up until birth to confirm normality. The median FBV was 1063.01 cm³ (range, 105.46-3262.4 cm³). We derived the following cubic equation showing the association between gestational age (GA) (from 16 to 36 GW) and FBV: GA (in days) = (3.84 × 10^-9 × FBV³) - (3.02 × 10^-5 × FBV²) + (0.103 × FBV) + 113.13. The mean ± SD difference between MRI-EFW and US-EFW was -27.58 ± 127.35 g for Baker's formula, -57.54 ± 130.19 g for Kacem's formula, 142.51 ± 120.12 g for Amgalan's formula, -113.43 ± 150.21 g for Debbi's formula and 12.65 ± 166.37 g for our model. The mean relative error on comparison of MRI-EFW according to Baker's formula with US-EFW was 29.33% at 16 GW and 2.98% at 36 GW, and the median was -0.74% (range, -34.36% to 29.33%) for the period 16-36 GW; according to Kacem's formula, the mean relative error was 24.10% at 16 GW and 1.87% at 36 GW, and the median was -3.31% (range, -36.14% to 24.10%) for the period 16-36 GW; according to Amgalan's formula, the mean relative error was 26.64% at 16 GW and 6.74% at 36 GW, and the median was 15.18% (range, -22.53% to 50.20%) for the period 16-36 GW; according to Debbi's formula, the mean relative error was 16.05% at 16 GW and - 0.63% at 36 GW, and the median was -7.25% (range, -42.26% to 16.54%) for the period 16-36 GW; and for our model, the mean relative error was 31.53% at 16 GW and 0.57% at 36 GW, and the median was 2.29% (range, -30.31% to 31.53%) for the period 16-36 GW. We provide a normative range for MRI-FBV from 16 to 36 GW, which can be used for clinical and research purposes. The strong association observed between MRI-FBV and US-EFW throughout pregnancy supports the potential for MRI-FBV to be used as a reliable alternative for EFW, particularly in clinical scenarios in which traditional US methods may be limited. Furthermore, the comparison of the four published MRI- or CT-based formulae with US-EFW indicated that our model had the strongest association with US-EFW. © 2025 International Society of Ultrasound in Obstetrics and Gynecology.