Accuracy of fetal lung volume assessed by three-dimensional sonography

Objective: To determine the accuracy and precision of prenatal three-dimensional (3D) ultrasound in estimating fetal lung volume using the rotational multiplanar technique (VOCAL™) by comparing it to postmortem volume measurements. Methods: Fetal lung volume was measured during 3D ultrasound examination using a rotational multiplanar technique in eight cases of congenital diaphragmatic hernia (CDH) (six left and two right-sided) and in 25 controls without pulmonary malformation, immediately before termination. Prenatal 3D sonographic estimates of fetal lung volume were compared with postmortem measurement of fetal lung volume achieved by water displacement. Results: The intraclass correlation coefficient of fetal lung volume estimated by 3D ultrasound and measured at postmortem examination was 0.95 in CDH cases and 0.99 in controls. Based on Bland-Altman analysis, the bias, precision and limits of agreement were, respectively, 0.35 cm³, 1.46 cm³ and between -2.51 and +3.21 cm³ in cases with CDH and 0.08 cm ³, 2.80 cm³ and between -5.41 and +5.57 cm³ in controls. The mean relative error of 3D ultrasound fetal lung volume measurement was -7.19% (from -42.70% to +18.11%) in CDH cases and -0.72% (from -30.25% to +19.22%) in controls, while the mean absolute error of 3D ultrasound fetal lung volume measurement was 1.40 (range, 0.71-2.52) cm³ and 2.12 (range, 0.05-4.98) cm³, respectively. Accuracy of 3D ultrasound for measuring fetal lung volumes was 84.86 (range, 57.30-99.48)% in cases with CDH and 91.38 (range, 69.75-99.45)% in controls. The mean intraobserver variability for lung volume estimated by 3D ultrasound was 0.28 cm³ in controls and 0.17 cm³ in CDH cases. Conclusion: Prenatal 3D ultrasound can estimate accurately fetal lung volume using the rotational multiplanar technique for volume measurements (VOCAL), even in fetuses with very small lungs, such as cases with isolated CDH.