Reproducibility of fetal brain measurements using 3D ultrasound

Document Type



Obstetrics and Gynaecology (East Africa)


Objectives: To establish the inter- and intra-operator reproducibility of key fetal brain structures throughout gestation using 3D ultrasound.

Methods: 3D volumes of the fetal head between 15 and 40 weeks of gestation were acquired prospectively using a standardized protocol as part of the INTERGROWTH-21st Project. For this reproducibility study a random sample of 90 volumes was used. The following measurements were taken: width of the posterior lateral ventricle in the transventricular plane; depth of the parieto-occipital fissure and Sylvian fissure in the transthalamic plane; and width of the cisterna magna and transcerebellar diameter in the transcerebellar plane. Operators worked in pairs. The first extracted the three planes of interest and took all measurements twice; this was repeated by the second operator. The effect of caliper replacement was also assessed. Mean differences and limits of agreement (LOA) were calculated using Bland-Altman plots.

Results: As there was a significant change with gestation in the transcerebellar diameter measurements, differences were expressed in percentage terms; the mean difference (95% LOA) was <1% (±12%). All other structures were compared in mm, and their mean differences (95% LOA) were ≤ 0.2mm (±2.8mm) (table 1). The greatest proportion of variability was due to caliper replacement.

Conclusions: Measurement of brain structures using standardized 3D fetal head volumes is highly reproducible. Caliper placement accounts for most of the observed variation among operators.


Ultrasound in Obstetrics and Gynecology