Regional growth patterns of fetal corpus callosum at 18–32 weeks' gestation assessed by three‐dimensional transvaginal neurosonography

Objectives To evaluate the normal development of the fetal corpus callosum (CC) between 18 and 32 weeks' gestation using three‐dimensional (3D) transvaginal neurosonography and to assess the effect of volume contrast imaging (VCI) on the measurement of fetal CC biometry. Methods This study included two cohorts of women with a singleton pregnancy: a prospective cohort attending the research clinic at Prince of Wales Hospital in Hong Kong SAR, China, between April 2019 and February 2025, at which 3D fetal brain volumes were acquired prospectively, and a retrospective cohort attending CRIFM Prenatal Medical Clinic in Osaka, Japan, between December 2018 and March 2025, from which archived 3D fetal brain volume datasets were retrieved. In both cohorts, 3D fetal brain volume data of the CC were acquired using transvaginal neurosonography between 18 and 32 weeks' gestation using the same acquisition protocol. Inclusion criteria were appropriately grown fetuses at 18 + 0 to 32 + 6 weeks' gestation, without suspected structural or chromosomal defects. In the midsagittal plane, using 3D volume multiplanar imaging, the CC length and total area were obtained and regional thickness at the genu, body and splenium were measured following a modified Garel's method. Additionally, the areas of five vertical CC segments defined by arithmetic fractions of the CC length following Witelson's classification (anterior one‐third, anterior midbody, posterior midbody, isthmus and splenium) were obtained. All fetal biometric measurements were performed using non‐VCI and VCI modes. Measurements obtained using non‐VCI and VCI modes were compared using the paired t ‐test. Generalized Additive Models for Location, Scale and Shape (GAMLSS) were used to explore the relationship between all biometric measurements of the fetal CC and gestational age. The final models were used to derive percentiles and construct reference curves. Intra‐ and interobserver variability were assessed using 30 randomly selected cases. Results A total of 956 fetuses at 18–32 weeks' gestation were included. Gestational age‐specific trajectories of fetal CC biometry were established, demonstrating distinct, non‐linear growth patterns in CC length, total area, genu, body and splenium thicknesses and the areas of five individual CC subdivisions from 18 to 32 weeks. Overall, the genu and body thicknesses increased sharply from 18 weeks to 23 weeks. Thereafter, genu thickness continued to increase gradually until 32 weeks, whereas body thickness plateaued before showing renewed growth at 31–32 weeks. The percentile curve for fetal CC total area demonstrated a steady increase from 18 to 32 weeks. Among the five subdivisions of the CC, the splenium demonstrated the most pronounced absolute area growth, reaching a median of approximately 0.40 cm 2 by 32 weeks, with markedly greater expansion than the midbody regions throughout the second to early third trimesters. Fetal CC biometric measurements were comparable between non‐VCI and VCI measurements. Overall, the biometric measurements demonstrated high reproducibility, with most showing good‐to‐excellent intra‐ and interobserver reliability, except for moderate interobserver reliability for genu and body thicknesses. Conclusions We have characterized region‐specific non‐linear growth patterns in fetal CC development from 18 to 32 weeks, with posterior‐predominant area expansion most pronounced in the splenium. The clinical significance of regional CC measurements remains uncertain, and the presented trajectories should be regarded as descriptions of normal developmental patterns rather than as reference charts for isolated clinical decision‐making. Longitudinal studies incorporating postnatal neurodevelopmental assessment will be required to determine their prognostic value.