Cerebral biometry in fetal magnetic resonance imaging: new reference data

Abstract

Objectives: To provide normal magnetic resonance imaging (MRI) reference biometric data of the fetal brain, to evaluate reproducibility and gender effect, to compare the two cerebral hemispheres and to compare MRI with ultrasonographic biometry, in a large cohort.

Methods: Normal cerebral fetal MRI examinations were collected prospectively and several parameters were measured: the supratentorial space (bone and cerebral fronto-occipital and biparietal (BPD) diameters), the length of the corpus callosum (LCC), the surface area, height and anteroposterior diameter of the vermis, the transverse cerebellar diameter (TCD) and the anteroposterior diameter of the pons. We evaluated the interobserver reproducibility of measurements and the possible gender effect on measurements of bone BPD, TCD and LCC. We compared right and left hemispheres, right and left atria and ultrasound and MRI measurements.

Results: The study included 589 fetuses, ranging from 26 to 40 weeks. Normal values (from 3(rd) to 97(th) percentile) are provided for each parameter. Interobserver agreement was excellent, with an intraclass correlation coefficient (ICC) > 0.75 for many parameters. The gender effect was evaluated in 372 cases and did not reveal any clinically meaningful difference. Comparison between the right and left cerebral hemispheres and between the right and left atria did not reveal any meaningful differences. Ultrasound and MRI measurements of BPD and TCD were compared in 94 cases and 48 cases, respectively, and the agreement was excellent (ICC = 0.85).

Conclusions: We present new reproducible reference charts for cerebral MRI biometry at 26-40 weeks' gestation, from a large cohort of fetuses.