Prenatal prediction of small-for-gestational age neonates using MR imaging: comparison with conventional 2D ultrasound

Abstract

Purpose: The purpose of this study is to evaluate the performance of estimating fetal weight (EFW) using magnetic resonance (MR) imaging as compared with two-dimensional (2D) ultrasound (US) in the prediction of small-for-gestational age neonates (SGA).Materials and methods: Written informed consent was obtained for this Ethical Committee approved study. Between March 2011 and May 2016, women with singleton pregnancies underwent US-EFW and MR-EFW within 48 h before delivery. US-EFW was based on Hadlock et al. and MR-EFW on the formula described by Backer et al. after planimetric measurement of the fetal body volume (FBV). Our outcome measure was performance in prediction of small-for-gestational age neonates by MR imaging versus US-EFW, using receiver-operating characteristic (ROC) curves.Results: Two hundred and seventy women were included in the study with 18 newborns (6.7%) of birthweight ≤10th, 12 (4.5%) ≤ 5th and 7 (2.6%) ≤ 3rd centile. The area under the ROC curve for prediction of birthweight ≤10th centile by prenatal MR imaging was significantly better than by US (difference between the AUROC = 0.060, p = .01; standard error = 0.023). Similarly, the area under the ROC curve for prediction of birthweight ≤5th centile by prenatal MR imaging was significantly better than by US (difference between the AUROC = 0.019, p = .03; standard error = 0.009). Finally, there was no significant difference between the areas under the ROC curve for the prediction of birthweight ≤3rd centile between the two imaging modalities (difference between the AUROC = 0.021, p = .13; standard error = 0.014).Conclusion: MR-EFW performed immediately prior to delivery predicts SGA neonates significantly better than US-EFW.