ISFB 2017 - The Second International Symposium on the Fetal Brain

In Vivo Placental Microstructure in the Growth-Restricted Fetus Using Magnetic Resonance Based Shape and Textural Analysis

Acute and chronic placental dysfunction has been associated with increased perinatal mortality and neurodevelopment disability in survivors.

Blog Article Figure
We intend to change trajectories to improve neurodevelopmental outcomes.

Ex vivo placental pathology reveals inflammatory and vascular abnormalities in placental-based fetal growth restriction (FGR) yet there is paucity of tools to assess in vivo placental health. Textural analysis is a quantitative image analysis method that has been used in biomedical imaging to distinguish pathologic from healthy tissue. The objective was to compare geometric and textural analysis measures of placental development in pregnancies complicated by FGR and uncomplicated, healthy pregnancies.

We recruited women with FGR and with normal pregnancies to undergo fetal MR imaging on 1.5T GE scanner using an 8-channel receiver coil.

Design/Methods

Cases were defined as FGR if the estimated fetal weight was < 10th % for age; cases with umbilical artery pulsatility index > 95th % for age OR with a cerebro-placental ratio < 1.0 were further categorized as FGR with Doppler anomalies. Any known or suspected genetic or infectious complications were excluded.

T2W-SSFSEs of the placental were acquired in the axial or coronal plane using 4mm slices. Placental segmentations were obtained using ITKSNAP software. Shape features were computed automatically based on an in-house 3D reconstruction method.

Textural features were calculated based on the previously segmented masks and 23 features were thus extracted for analysis. Linear mixed models with robust variance estimate were used to evaluate the association between textural properties in FGR vs. controls, permitting multiple measurements on the same patient. Outcome variables were log transformed prior to analysis and all models adjusted for gestational age at the time of the assessment.

Results

We studied 66 women with a mean age (GA) of 31 weeks; 21 underwent serial scans for a total of 87 images (45 controls, 42 FGR). Thirteen features were significantly different in FGR compared to controls; collectively, these features demonstrate smaller placentas with increased symmetry, homogeneity and finer textures in FGR compared to controls; sub-group analysis revealed further differences between FGR cases with and without Doppler abnormalities.