In recent years there have been many studies indicating that multiple cortical features, extracted at each surface vertex, are promising in the detection of various neurodevelopmental and neurodegenerative diseases. However, with limited datasets, it is challenging to train stable classifiers with such high-dimensional surface data. This necessitates a feature reduction that is commonly accomplished via regional volumetric morphometry from standard brain atlases. However, current regional summaries are not specific to the given age or pathology that is studied, which runs the risk of losing relevant information that can be critical in the classification process. To solve this issue, this paper proposes a novel data-driven approach by extending convolutional neural networks (CNN) for use on non-Euclidean manifolds such as cortical surfaces. The proposed network learns the most powerful features and brain regions from the extracted large dimensional feature space; thus creating a new feature space in which the dimensionality is reduced and feature distributions are better separated. We demonstrate the usability of the proposed surface-CNN framework in an example study classifying Alzheimers disease patients versus normal controls. The high performance in the cross-validation diagnostic results shows the potential of our proposed prediction system.
A broadband MIMO antenna, using
Introduction The prenatal period is characterized by immense fetal neuronal growth. Such rapid growth can increase fetal susceptibility to prenatal environmental insults (Barker, 1998). A promising prenatal process that may alter fetal development is maternal prenatal sleep quality. Poor prenatal sleep quality is a public health concern affecting approximately 78% of pregnant individuals (Lucena et al., 2018). In rodents, maternal sleep deprivation across gestation predicts offspring hippocampal neurogenesis, with pups exposed to sleep deprivation early and late in pregnancy exhibiting more anxiety and depression-like behaviors (Peng et al., 2015). In humans, poor sleep quality in other developmental stages predicts hippocampi and amygdalae changes (Marshall & Born, 2007; Saghir et al., 2018). However, the relation between prenatal sleep quality and offspring brain development in humans remains poorly understood. The present study examined associations between maternal sleep quality in early, mid, and late pregnancy, and newborn hippocampal and amygdala volume, regions implicated in memory and emotion. Methods Pregnant individuals (N=94; Mage=30.5; SDage=5.3) reported on sleep quality three times during pregnancy. Newborn (Mageinweeks=5.1; SDageinweeks=2.7) hippocampi and amygdalae volumes were assessed during an unsedated sleep cycle using magnetic resonance imaging (MRI). Tissue segmentation was collected using a multiatlas iterative algorithm that individually segmented the regions of interest and subsequently combined T1- and T2-weighted high-resolution images (See neonate multiatlas at https://www.nitrc.org/projects/unc_brain_atlas/). Bivariate correlations examined the association between prenatal sleep quality and hippocampus and amygdala volume. Partial correlations examined these associations in the presence of significant cofounding variables including intracranial volume, body weight percentile, sex, and postconceptional age. Results Partial correlations revealed that poor maternal sleep quality early in pregnancy predicted larger newborn bilateral hippocampal volume (all rs<.25; ps<.038). Associations with sleep later in gestation persisted for the right hippocampus (all rs<.25; ps<.038). Prenatal maternal sleep quality did not significantly predict newborn amygdala volume (all rs<-.06; ps>.58). Conclusion This study provides novel evidence linking prenatal sleep quality and newborn hippocampal volume in humans, suggesting the presence of an intergenerational link between prenatal sleep health and offspring well-being. Support (If Any) Support (if any): NIMH R01MH109662, NHLBI R01HL155744, and diversity training supplement for 1st author; F32MH125572 for 2nd author.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.