Prenatal delta-9-tetrahydrocannabinol exposure alters fetal neurodevelopment in rhesus macaques

Prenatal cannabis use is associated with adverse offspring neurodevelopmental outcomes, however the underlying mechanisms are relatively unknown. We sought to determine the impact of chronic delta-9-tetrahydrocannabinol (THC) exposure on fetal neurodevelopment in a rhesus macaque model using advanced imaging combined with molecular and tissue studies. Animals were divided into two groups, control (n = 5) and THC-exposed (n = 5), which received a daily THC edible pre-conception and throughout pregnancy. Fetal T2-weighted MRI was performed at gestational days 85 (G85), G110, G135 and G155 to assess volumetric brain development. At G155, animals underwent cesarean delivery with collection of fetal cerebrospinal fluid (CSF) for microRNA (miRNA) studies and fetal tissue for histologic analysis. THC exposure was associated with significant age by sex interactions in brain growth, and differences in fetal brain histology suggestive of brain dysregulation. Two extracellular vesicle associated-miRNAs were identified in THC-exposed fetal CSF; pathway analysis suggests that these miRNAs are associated with dysregulated axonal guidance and netrin signaling. This data is indicative of subtle molecular changes consistent with the observed histological data, suggesting a potential role for fetal miRNA regulation by THC. Further studies are needed to determine whether these adverse findings correlate with long-term offspring neurodevelopmental health.

Subcortical Segmentation of the Fetal Brain in 3D Ultrasound using Deep Learning

The quantification of subcortical volume development from 3D fetal ultrasound can provide important diagnostic information during pregnancy monitoring. However, manual segmentation of subcortical structures in ultrasound volumes is time-consuming and challenging due to low soft tissue contrast, speckle and shadowing artifacts. For this reason, we developed a convolutional neural network (CNN) for the automated segmentation of the choroid plexus (CP), lateral posterior ventricle horns (LPVH), cavum septum pellucidum et vergae (CSPV), and cerebellum (CB) from 3D ultrasound. As ground-truth labels are scarce and expensive to obtain, we applied few-shot learning, in which only a small number of manual annotations (n = 9) are used to train a CNN. We compared training a CNN with only a few individually annotated volumes versus many weakly labelled volumes obtained from atlas-based segmentations. This showed that segmentation performance close to intra-observer variability can be obtained with only a handful of manual annotations. Finally, the trained models were applied to a large number (n = 278) of ultrasound image volumes of a diverse, healthy population, obtaining novel US-specific growth curves of the respective structures during the second trimester of gestation.