In our study we enrolled 40 pregnant women [gestational age (GA) 20-38 weeks] who were referred to our department to undergo prenatal MRI examination after obstetric US, from 2017 to 2019.
The study inclusion criteria were the absence of fetal and maternal pathology and good image quality. The final study cohort was composed of 27 normal pregnancies fulfilling the inclusion study criteria.
Written informed consent was obtained by all pregnant women prior to fetal MR imaging. No sedation or contrast agent was administered. Prenatal MRI studies were performed on pregnant woman in the supine or left lateral position, using a 1.5 T superconducting unit (Siemens Magnetom Avanto) and two body matrix coils were placed on the maternal abdomen.
Standard fetal body MRI protocol included T2-HASTE (Half Fourier Single-shot Turbo Spin Echo) Weighted Imaging (WI), T1 WI with and without Fat Suppression (FS), and TrueFISP WI.
The IVIM protocol included a DW Echo-Planar Imaging (EPI) sequence on axial plane of fetal body, with TR/TE = 4000ms/79ms; bandwidth 1628 Hz/px; matrix-size 192x192; N° of slices 30; in plane resolution 2.0x2.0 mm2; slice thickness 4 mm.
Diffusion encoding gradients were applied along 3 no-coplanar directions using 10 different B values (0, 10, 30, 50, 75, 100, 150, 400, 700, 1000 s/mm2).
The total acquisition time of the protocol was about 6 min.
Image quality selection was made by two radiologists with fourteen years and five years of experience in prenatal MR. Fig. 1 Fig. 2
8 MR acquisition volumes of fetal kidneys were excluded from the measurement due to movement artifacts.
DICOM images of DWI acquisitions were elaborated offline with a prototype software named 'Siemens MR Body Diffusion Toolbox' to obtain 3 parameters: perfusion fraction f, pseudo-diffusion D* and diffusion coefficient D.
f quantifies the perfusion fraction of water molecules that perfuse in microcapillaries with D* rate, whereas D quantifies the real diffusion of water molecules in the extracellular space and it is related to tissue microstructure.
Signal intensity was averaged in each selected ROI, and data were fitted to the function:
S(b)/S(0) = f exp (−b D*) + (1 – f) exp (−b D)
where S(b) is the diffusion-weighted signal and S(0) is the signal at b=0s/mm2.
The differential signal intensity attenuation at different b values was described by a bi-exponential curve: the first part of the curve (low b-values) describes the perfusion compartment (f and D*), while second part of the curve (higher b-values) describes the real diffusion of water molecules in the extracellular space (D). Fig. 3
It means that signal attenuation at low b values depends not only on water diffusion in tissues but alsoì on microcirculation in the capillary network.
With the aim to perform an accurate IVIM analysis, for each foetus, two bilateral ROIs were manually placed on fetal kidney and lung parenchyma, excluding the hili.
Mean values of fraction of perfusion f, Pseudo-Diffusion Coefficient D* and Diffusion Coefficient D were obtained.
We have investigated how Ivim parameters change during gestation.
Differences between ROIs f, D, D* mean values were assessed with Analysis of Variance (ANOVA) test with Bonferroni correction. Pearson test with Bonferroni correction was performed to investigate the correlation between IVIM parameters (D, D*, f) and Gestational Age (GA). All statistical analysis was performed using SPSS Statistics 20 (IBM SPSS, Inc. Chicago, IL).