MR volumetry of brain and CSF in fetuses referred for ventriculomegaly

Abstract

Objective: The purpose of this study was to validate the method of performing fetal brain volumetry. In particular, our objectives were to assess which imaging plane is most reproducible for the performance of brain volumetry measurements and to ascertain inter- and intraobserver variability in determining brain volume in fetuses referred for ventriculomegaly (VM).

Subjects and methods: In this prospective study, 50 consecutive fetuses at 17-37 weeks of gestational age referred for MRI for VM underwent fast spin-echo T2-weighted imaging. Supratentorial brain parenchyma, lateral ventricles, and extraaxial and cerebellar volumetric measurements were manually obtained in three planes by three radiologists. Inter- and intraobserver variability were assessed. The relationship between volumes and gestational age, and lateral ventricular diameter were assessed.

Results: Volumes increased with gestational age. The presence of VM correlated with increased lateral ventricle diameter. The effect of imaging plane was negligible. Inter- and intraobserver variability were low.

Conclusion: Supratentorial parenchyma and lateral ventricular volumes can be reliably measured on fetal MRI, and imaging plane was not an important factor in measurement. Further studies are needed to correlate these indexes with long-term postnatal outcomes.

Figures

Figure 1

Coronal views of fetus at 31 weeks demonstrating tracings of the supratentorial cortex (A), ventricles (B), and cerebellum (C). The supratentorial parenchymal volume was determined by subtracting the values for the ventricles from that of the cortex.

Figure 2

Axial image at 31 weeks demonstrating the tracing of extra-axial CSF

Figure 3

Volume of supratentorial parenchyma was directly related to fetal GA as determined by ultrasound. A similar pattern was observed with GA determined by dates, BPD, and HC (see Table 2).

Figure 4

Volume of ventricles was directly related to fetal GA as determined by ultrasound but increased less rapidly than parenchymal volume (Fig. 3). A similar pattern was observed with GA determined by dates, BPD and HC (see Table 2).

Figure 5

Volume of cerebellum (including vermis) was directly related to fetal GA as determined by ultrasound (l.) and brain size as represented by BPD (r.). Rate of increase was comparable to that of parenchyma (Fig. 3) and varied significantly by plane of measurement. Similar patterns were observed with GA determined by dates or by brain size represented by HC (see Table 2).

Figure 6

Parenchymal volume measured by MR, grouped by prenatal diagnosis and preliminary neonatal outcome. normal outcome; + − − − − mild isolated VM 10-15 mm prenatal diagnosis with either normal outcome or isolated mild VM; ▾ ·········· all other VM cases. For illustration each point represents an average over all 3 planes and 2 raters; the analysis was performed on individual data points.

Figure 7

Ventricular volume measured by MR, grouped by prenatal diagnosis and preliminary neonatal outcome. normal outcome; + − − − − mild isolated VM 10-15 mm prenatal diagnosis with either normal outcome or isolated mild VM; ▾ ·········· all other VM cases. For illustration each point represents an average over all 3 planes and 2 raters; the analysis was performed on individual data points.