Showing papers by "Maria Deprez published in 2021"

Antenatal thymus volumes in fetuses that delivered <32 weeks' gestation: An MRI pilot study.

Abstract: INTRODUCTION Infection and inflammation have been implicated in the etiology and subsequent morbidity associated with preterm birth. At present, there are no tests to assess for fetal compartment infection. The thymus, a gland integral in the fetal immune system, has been shown to involute in animal models of antenatal infection, but its response in human fetuses has not been studied. This study aims: (a) to generate magnetic resonance imaging (MRI) -derived fetal thymus volumes standardized for fetal weight; (b) to compare standardized thymus volumes from fetuses that delivered before 32 weeks of gestation with fetuses that subsequently deliver at term; (c) to assess thymus size as a predictor of preterm birth; and (d) to correlate the presence of chorioamnionitis and funisitis at delivery with thymic volumes in utero in fetuses that subsequently deliver preterm. MATERIAL AND METHODS Women at high-risk of preterm birth at 20-32 weeks of gestation were recruited. A control group was obtained from existing data sets acquired as part of three research studies. A fetal MRI was performed on a 1.5T or 3T MRI scanner: T2 weighted images were obtained of the entire uterine content and specifically the fetal thorax. A slice-to-volume registration method was used for reconstruction of three-dimensional images of the thorax. Thymus segmentations were performed manually. Body volumes were calculated by manual segmentation and thymus:body volume ratios were generated. Comparison of groups was performed using multiple regression analysis. Normal ranges were created for thymus volume and thymus:body volume ratios using the control data. Receiver operating curves (ROC) curves were generated for thymus:body volume ratio and gestation-adjusted thymus volume centiles as predictors of preterm birth. Placental histology was analyzed where available from pregnancies that delivered very preterm and the presence of chorioamnionitis/funisitis was noted. RESULTS Normative ranges were created for thymus volume, and thymus volume was standardized for fetal size from fetuses that subsequently delivered at term, but were imaged at 20-32 weeks of gestation. Image data sets from 16 women that delivered 37 weeks were included. Mean gestation at MRI of the study group was 28+4 weeks (SD 3.2) and for the control group was 25+5 weeks (SD 2.4). Both absolute fetal thymus volumes and thymus:body volume ratios were smaller in fetuses that delivered preterm (P < .001). Of the 16 fetuses that delivered preterm, 13 had placental histology, 11 had chorioamnionitis, and 9 had funisitis. The strongest predictors of prematurity were the thymus volume Z-score and thymus:body volume ratio Z-score (ROC areas 0.915 and 0.870, respectively). CONCLUSIONS We have produced MRI-derived normal ranges for fetal thymus and thymus:body volume ratios between 20 and 32 weeks of gestation. Fetuses that deliver very preterm had reduced thymus volumes when standardized for fetal size. A reduced thymus volume was also a predictor of spontaneous preterm delivery. Thymus volume may be a suitable marker of the fetal inflammatory response, although further work is needed to assess this, increasing the sample size to correlate the extent of chorioamnionitis with thymus size.

Harmonized Segmentation of Neonatal Brain MRI.

Abstract: Deep learning based medical image segmentation has shown great potential in becoming a key part of the clinical analysis pipeline. However, many of these models rely on the assumption that the train and test data come from the same distribution. This means that such methods cannot guarantee high quality predictions when the source and target domains are dissimilar due to different acquisition protocols, or biases in patient cohorts. Recently, unsupervised domain adaptation techniques have shown great potential in alleviating this problem by minimizing the shift between the source and target distributions, without requiring the use of labelled data in the target domain. In this work, we aim to predict tissue segmentation maps on T2-weighted magnetic resonance imaging data of an unseen preterm-born neonatal population, which has both different acquisition parameters and population bias when compared to our training data. We achieve this by investigating two unsupervised domain adaptation techniques with the objective of finding the best solution for our problem. We compare the two methods with a baseline fully-supervised segmentation network and report our results in terms of Dice scores obtained on our ground truth test dataset. Moreover, we analyse tissue volumes and cortical thickness measures of the harmonised data on a subset of the population matched for gestational age at birth and postmenstrual age at scan. Finally, we demonstrate the applicability of the harmonised cortical gray matter maps with an analysis comparing term and preterm-born neonates and a proof-of-principle investigation of the association between cortical thickness and a language outcome measure.

Multi-channel 4D parametrized Atlas of Macro- and Microstructural Neonatal Brain Development

Abstract: Structural (also known as anatomical) and diffusion MRI provide complimentary anatomical and microstructural characterization of early brain maturation. However, the existing models of the developing brain in time include only either structural or diffusion MRI channels. Furthermore, there is a lack of tools for combined analysis of structural and diffusion MRI in the same reference space. In this work, we propose a methodology to generate a multi-channel (MC) continuous spatio-temporal parametrized atlas of the brain development that combines multiple MRI-derived parameters in the same anatomical space during 37-44 weeks of postmenstrual age range. We co-align structural and diffusion MRI of 170 normal term subjects from the developing Human Connectomme Project using MC registration driven by both T2-weighted and orientation distribution functions channels and fit the Gompertz model to the signals and spatial transformations in time. The resulting atlas consists of 14 spatio-temporal microstructural indices and two parcellation maps delineating white matter tracts and neonatal transient structures. In order to demonstrate applicability of the atlas for quantitative region-specific studies, a comparison analysis of 140 term and 40 preterm subjects scanned at the term-equivalent age is performed using different MRI-derived microstructural indices in the atlas reference space for multiple white matter regions, including the transient compartments. The atlas and software will be available after publication of the article.

Automated 3D reconstruction of the fetal thorax in the standard atlas space from motion-corrupted MRI stacks for 21-36 weeks GA range

Abstract: Slice-to-volume registration (SVR) methods allow reconstruction of high-resolution 3D images from multiple motion-corrupted stacks. SVR-based pipelines have been increas- ingly used for motion correction for fetal MRI since they allow more informed and de- tailed diagnosis of brain and body anomalies including congenital heart defects (Lloyd et al., 2019). Recently, fully automated rigid SVR reconstruction of the fetal brain in the atlas space was achieved in (Salehi et al., 2019) that used segmentation and pose es- timation convolutional neural networks (CNNs). However, these CNN-based methods have not yet been applied to the fetal body region. Meanwhile, the existing rigid and deformable SVR (DSVR) solutions (Uus et al., 2020) for the fetal trunk region are limited by the requirement of manual input as well the narrow capture range of the classical gradient descent based registration methods that cannot resolve severe fetal motion fre- quently occurring at the early gestational age (GA). Furthermore, in our experience, the conventional 2D slice-wise CNN-based brain masking solutions are reportedly prone to errors that require manual corrections when applied on a wide range of acquisition protocols or abnormal cases in clinical setting. In this work, we propose a fully automated pipeline for reconstruction of the fetal thorax region for 21-36 weeks GA range MRI datasets. It includes 3D CNN-based intra-uterine localisation of the fetal trunk and landmark-guided pose estimation steps that allow automated DSVR reconstruction in the standard radiological space irrespec- tive of the fetal body position or the regional stack coverage. The additional step for generation of the common template space and rejection of outliers provides the means for automated exclusion of stacks affected by low image quality or extreme motion. The pipeline was evaluated on a series of experiments including fetal MRI datasets and simulated rotation motion. Furthermore, we performed a qualitative assessment of the image reconstruction quality in terms of the definition of vascular structures on 100 early (median 23.14 weeks) and late (median 31.79 weeks) GA group MRI datasets covering 21 to 36 weeks GA range.

Spatio-Temporal Atlas of Normal Fetal Craniofacial Feature Development and CNN-Based Ocular Biometry for Motion-Corrected Fetal MRI

Abstract: Motion-corrected fetal magnetic resonance imaging (MRI) is widely employed in large-scale fetal brain studies. However, the current processing pipelines and spatio-temporal atlases tend to omit craniofacial structures, which are known to be linked to genetic syndromes. In this work, we present the first spatio-temporal atlas of the fetal head that includes craniofacial features and covers 21 to 36 weeks gestational age range. Additionally, we propose a fully automated pipeline for fetal ocular biometry based on a 3D convolutional neural network (CNN). The extracted biometric indices are used for the growth trajectory analysis of changes in ocular metrics for 253 normal fetal subjects from the developing human connectome project (dHCP).

Multi-channel 4D parametrized Atlas of Macro- and Microstructural Neonatal Brain Development

Abstract: Structural and diffusion MRI provide complimentary anatomical and microstructural characterization of early brain maturation. The existing models of the developing brain in time include only either structural or diffusion channels. Furthermore, there is a lack of tools for combined analysis of structural and diffusion MRI in the same reference space. In this work we propose methodology to generate multi-channel (MC) continuous spatio-temporal parametrized atlas of brain development based on MC registration driven by both T2-weighted and orientation distribution functions (ODF) channels along with the Gompertz model (GM) fitting of the signals and spatial transformations in time. We construct a 4D MC atlas of neonatal brain development during 38 to 44 week PMA range from 170 normal term subjects from developing Human Connectomme Project. The resulting atlas consists of fourteen spatio-temporal microstructural indices and two parcellation maps delineating white matter tracts and neonatal transient structures. We demonstrate applicability of the atlas for quantitative region-specific comparison of 140 term and 40 preterm subjects scanned at the term-equivalent age. We show multi-parametric microstructural differences in multiple white matter regions, including the transient compartments. The atlas and software will be available after publication of the article.

3D UNet with GAN discriminator for robust localisation of the fetal brain and trunk in MRI with partial coverage of the fetal body

Abstract: In fetal MRI, automated localisation of the fetal brain or trunk is a prerequisite for motion correction methods. However, the existing CNN-based solutions are prone to errors and may require manual editing. In this work, we propose to combine a multi-label 3D UNet with a GAN discriminator for localisation of both fetal brain and trunk in fetal MRI stacks. The proposed method is robust for datasets with both full and partial coverage of the fetal body.

Uncertainty-Aware Deep Learning Based Deformable Registration

Abstract: We introduce an uncertainty-aware deep learning deformable image registration solution for magnetic resonance imaging multi-channel data. In our proposed framework, the contributions of structural and microstructural data to the displacement field are weighted with spatially varying certainty maps. We produce certainty maps by employing a conditional variational autoencoder image registration network, which enables us to generate uncertainty maps in the deformation field itself. Our approach is quantitatively evaluated on pairwise registrations of 36 neonates to a standard structural and/or microstructural template, and compared with models trained on either single modality, or both modalities together. Our results show that by incorporating uncertainty while fusing the two modalities, we achieve superior alignment in cortical gray matter and white matter regions, while also achieving a good alignment of the white matter tracts. In addition, for each of our trained models, we show examples of average uncertainty maps calculated for 10 neonates scanned at 40 weeks post-menstrual age.

Harmonised segmentation of neonatal brain MRI

Abstract: Deep learning based medical image segmentation has shown great potential in becoming a key part of the clinical analysis pipeline. However, many of these models rely on the assumption that the train and test data come from the same distribution. This means that such methods cannot guarantee high quality predictions when the source and target domains are dissimilar due to different acquisition protocols, or biases in patient cohorts. Recently, unsupervised domain adaptation (DA) techniques have shown great potential in alleviating this problem by minimizing the shift between the source and target distributions, without requiring the use of labelled data in the target domain. In this work, we aim to predict tissue segmentation maps on T2-weighted (T2w) magnetic resonance imaging (MRI) data of an unseen preterm-born neonatal population, which has both different acquisition parameters and population bias when compared to our training data. We achieve this by investigating two unsupervised DA techniques with the objective of finding the best solution for our problem. We compare the two methods with a baseline fully-supervised segmentation network and report our results in terms of Dice scores obtained on our ground truth test dataset. Moreover, we analyse tissue volumes and cortical thickness (CT) measures of the harmonised data on a subset of the population matched for gestational age (GA) at birth and postmenstrual age (PMA) at scan. Finally, we demonstrate the applicability of the harmonised cortical gray matter maps with an analysis comparing term and preterm-born neonates and a proof-of-principle investigation of the association between CT and a language outcome measure.

Fetal MRI by robust deep generative prior reconstruction and diffeomorphic registration: application to gestational age prediction.

Abstract: Magnetic resonance imaging of whole fetal body and placenta is limited by different sources of motion affecting the womb. Usual scanning techniques employ single-shot multi-slice sequences where anatomical information in different slices may be subject to different deformations, contrast variations or artifacts. Volumetric reconstruction formulations have been proposed to correct for these factors, but they must accommodate a non-homogeneous and non-isotropic sampling, so regularization becomes necessary. Thus, in this paper we propose a deep generative prior for robust volumetric reconstructions integrated with a diffeomorphic volume to slice registration method. Experiments are performed to validate our contributions and compare with a state of the art method in a cohort of $72$ fetal datasets in the range of $20-36$ weeks gestational age. Results suggest improved image resolution and more accurate prediction of gestational age at scan when comparing to a state of the art reconstruction method. In addition, gestational age prediction results from our volumetric reconstructions compare favourably with existing brain-based approaches, with boosted accuracy when integrating information of organs other than the brain. Namely, a mean absolute error of $0.618$ weeks ($R^2=0.958$) is achieved when combining fetal brain and trunk information.