Automatic segmentation of MR images of the developing newborn brain

Longitudinal trajectories of brain development from infancy to school age and their relationship to literacy development

Reading is one of the most complex skills that we utilize daily, and it involves the early development and interaction of various lower-level subskills, including phonological processing and oral language. These subskills recruit brain structures, which begin to develop long before the skill manifests and exhibit rapid development during infancy. However, how longitudinal trajectories of early brain development in these structures support long-term acquisition of literacy subskills and subsequent reading is unclear. Children underwent structural and diffusion MRI scanning at multiple timepoints between infancy and second grade and were tested for literacy subskills in preschool and decoding and word reading in early elementary school. We developed and implemented a reproducible pipeline to generate longitudinal trajectories of early brain development to examine associations between these trajectories and literacy (sub)skills. Furthermore, we examined whether familial risk of reading difficulty and children's home literacy environments, two common literacy-related covariates, influenced those trajectories. Results showed that individual differences in curve features (e.g., intercepts and slopes) for longitudinal trajectories of volumetric, surface-based, and white matter organization measures were linked directly to phonological processing and indirectly to first-grade decoding and word reading skills via phonological processing. Altogether, these findings suggest that the brain bases of phonological processing, previously identified as the strongest behavioral predictor of reading and decoding skills, may already begin to develop by birth but undergo further refinement between infancy and preschool. The present study underscores the importance of considering academic skill acquisition from the very beginning of life.

Automatic Brain Segmentation for PET/MR Dual-Modal Images Through a Cross-Fusion Mechanism

The precise segmentation of different brain regions and tissues is usually a prerequisite for the detection and diagnosis of various neurological disorders in neuroscience. Considering the abundance of functional and structural dual-modality information for positron emission tomography/magnetic resonance (PET/MR) images, we propose a novel 3D whole-brain segmentation network with a cross-fusion mechanism introduced to obtain 45 brain regions. Specifically, the network processes PET and MR images simultaneously, employing UX-Net and a cross-fusion block for feature extraction and fusion in the encoder. We test our method by comparing it with other deep learning-based methods, including 3DUXNET, SwinUNETR, UNETR, nnFormer, UNet3D, NestedUNet, ResUNet, and VNet. The experimental results demonstrate that the proposed method achieves better segmentation performance in terms of both visual and quantitative evaluation metrics and achieves more precise segmentation in three views while preserving fine details. In particular, the proposed method achieves superior quantitative results, with a Dice coefficient of 85.73% 0.01%, a Jaccard index of 76.68% 0.02%, a sensitivity of 85.00% 0.01%, a precision of 83.26% 0.03% and a Hausdorff distance (HD) of 4.4885 14.85%. Moreover, the distribution and correlation of the SUV in the volume of interest (VOI) are also evaluated (PCC > 0.9), indicating consistency with the ground truth and the superiority of the proposed method. In future work, we will utilize our whole-brain segmentation method in clinical practice to assist doctors in accurately diagnosing and treating brain diseases.

Structure-Aware Brain Tissue Segmentation for Isointense Infant MRI Data Using Multi-Phase Multi-Scale Assistance Network

Accurate and automatic brain tissue segmentation is crucial for tracking brain development and diagnosing brain disorders. However, due to inherently ongoing myelination and maturation during the first postnatal year, the intensity distributions of gray matter and white matter in the infant brain MRI at the age of around 6 months old (a.k.a. isointense phase) are highly overlapped, which makes tissue segmentation very challenging, even for experts. To address this issue, in this study, we propose a multi-phase multi-scale assistance segmentation framework, which comprises a structure-preserved generative adversarial network (SPGAN) and a multi-phase multi-scale assisted segmentation network (MASN). SPGAN bi-directionally synthesizes isointense and adult-like data. The synthetic isointense data essentially augment the training dataset, combined with high-quality annotations transferred from its adult-like counterpart. By contrast, the synthetic adult-like data offers clear tissue structures and is concatenated with isointense data to serve as the input of MASN. In particular, MASN is designed with two-branch networks, which simultaneously segment tissues with two phases (isointense and adult-like) and two scales by also preserving their correspondences. We further propose a boundary refinement module to extract maximum gradients from local feature maps to indicate tissue boundaries, prompting MASN to focus more on boundaries where segmentation errors are prone to occur. Extensive experiments on the National Database for Autism Research and Baby Connectome Project datasets quantitatively and qualitatively demonstrate the superiority of our proposed framework compared with seven state-of-the-art methods.

The Developing Human Connectome Project: A fast deep learning-based pipeline for neonatal cortical surface reconstruction

The Developing Human Connectome Project (dHCP) aims to explore developmental patterns of the human brain during the perinatal period. An automated processing pipeline has been developed to extract high-quality cortical surfaces from structural brain magnetic resonance (MR) images for the dHCP neonatal dataset. However, the current implementation of the pipeline requires more than 6.5 h to process a single MRI scan, making it expensive for large-scale neuroimaging studies. In this paper, we propose a fast deep learning (DL) based pipeline for dHCP neonatal cortical surface reconstruction, incorporating DL-based brain extraction, cortical surface reconstruction and spherical projection, as well as GPU-accelerated cortical surface inflation and cortical feature estimation. We introduce a multiscale deformation network to learn diffeomorphic cortical surface reconstruction end-to-end from T2-weighted brain MRI. A fast unsupervised spherical mapping approach is integrated to minimize metric distortions between cortical surfaces and projected spheres. The entire workflow of our DL-based dHCP pipeline completes within only 24 s on a modern GPU, which is nearly 1000 times faster than the original dHCP pipeline. The qualitative assessment demonstrates that for 82.5% of the test samples, the cortical surfaces reconstructed by our DL-based pipeline achieve superior (54.2%) or equal (28.3%) surface quality compared to the original dHCP pipeline.

BIBSNet: A Deep Learning Baby Image Brain Segmentation Network for MRI Scans

Objectives

Brain segmentation of infant magnetic resonance (MR) images is vitally important for studying typical and atypical brain development. The infant brain undergoes many changes throughout the first years of postnatal life, making tissue segmentation difficult for most existing algorithms. Here we introduce a deep neural network BIBSNet ( B aby and I nfant B rain S egmentation Neural Net work), an open-source, community-driven model for robust and generalizable brain segmentation leveraging data augmentation and a large sample size of manually annotated images.

Experimental Design

Included in model training and testing were MR brain images from 90 participants with an age range of 0-8 months (median age 4.6 months). Using the BOBs repository of manually annotated real images along with synthetic segmentation images produced using SynthSeg, the model was trained using a 10-fold procedure. Model performance of segmentations was assessed by comparing BIBSNet, joint label fusion (JLF) inferred segmentation to ground truth segmentations using Dice Similarity Coefficient (DSC). Additionally, MR data along with the FreeSurfer compatible segmentations were processed with the DCAN labs infant-ABCD-BIDS processing pipeline from ground truth, JLF, and BIBSNet to further assess model performance on derivative data, including cortical thickness, resting state connectivity and brain region volumes.

Principal Observations

BIBSNet segmentations outperforms JLF across all regions based on DSC comparisons. Additionally, with processed derived metrics, BIBSNet segmentations outperforms JLF segmentations across nearly all metrics.

Conclusions

BIBSNet segmentation shows marked improvement over JLF across all age groups analyzed. The BIBSNet model is 600x faster compared to JLF, produces FreeSurfer-compatible segmentation labels, and can be easily included in other processing pipelines. BIBSNet provides a viable alternative for segmenting the brain in the earliest stages of development.

Subject-specific atlas for automatic brain tissue segmentation of neonatal magnetic resonance images

Developing advanced systems for 3D brain tissue segmentation from neonatal magnetic resonance (MR) images is vital for newborn structural analysis. However, automatic segmentation of neonatal brain tissues is challenging due to smaller head size and inverted T1/T2 tissue contrast compared to adults. In this work, a subject-specific atlas based technique is presented for segmentation of gray matter (GM), white matter (WM), and cerebrospinal fluid (CSF) from neonatal MR images. It involves atlas selection, subject-specific atlas creation using random forest (RF) classifier, and brain tissue segmentation using the expectation maximization-Markov random field (EM-MRF) method. To increase the segmentation accuracy, different tissue intensity- and gradient-based features were used. Evaluation on 40 neonatal MR images (gestational age of 37-44 weeks) demonstrated an overall accuracy of 94.3% and an average Dice similarity coefficient (DSC) of 0.945 (GM), 0.947 (WM), and 0.912 (CSF). Compared to multi-atlas segmentation methods like SEGMA and EM-MRF with multiple atlases, our method improved accuracy by up to 4%, particularly in complex tissue regions. Our proposed method allows accurate brain tissue segmentation, a crucial step in brain magnetic resonance imaging (MRI) applications including brain surface reconstruction and realistic head model creation in neonates.

VINNA for neonates: Orientation independence through latent augmentations

A robust, fast, and accurate segmentation of neonatal brain images is highly desired to better understand and detect changes during development and disease, specifically considering the rise in imaging studies for this cohort. Yet, the limited availability of ground truth datasets, lack of standardized acquisition protocols, and wide variations of head positioning in the scanner pose challenges for method development. A few automated image analysis pipelines exist for newborn brain Magnetic Resonance Image (MRI) segmentation, but they often rely on time-consuming non-linear spatial registration procedures and require resampling to a common resolution, subject to loss of information due to interpolation and down-sampling. Without registration and image resampling, variations with respect to head positions and voxel resolutions have to be addressed differently. In deep learning, external augmentations such as rotation, translation, and scaling are traditionally used to artificially expand the representation of spatial variability, which subsequently increases both the training dataset size and robustness. However, these transformations in the image space still require resampling, reducing accuracy specifically in the context of label interpolation. We recently introduced the concept of resolution-independence with the Voxel-size Independent Neural Network framework, VINN. Here, we extend this concept by additionally shifting all rigid-transforms into the network architecture with a four degree of freedom (4-DOF) transform module, enabling resolution-aware internal augmentations (VINNA) for deep learning. In this work, we show that VINNA (i) significantly outperforms state-of-the-art external augmentation approaches, (ii) effectively addresses the head variations present specifically in newborn datasets, and (iii) retains high segmentation accuracy across a range of resolutions (0.5-1.0 mm). Furthermore, the 4-DOF transform module together with internal augmentations is a powerful, general approach to implement spatial augmentation without requiring image or label interpolation. The specific network application to newborns will be made publicly available as VINNA4neonates.

Deep learning techniques for isointense infant brain tissue segmentation: a systematic literature review

Introduction

To improve comprehension of initial brain growth in wellness along with sickness, it is essential to precisely segment child brain magnetic resonance imaging (MRI) into white matter (WM) and gray matter (GM), along with cerebrospinal fluid (CSF). Nonetheless, in the isointense phase (6-8 months of age), the inborn myelination and development activities, WM along with GM display alike stages of intensity in both T1-weighted and T2-weighted MRI, making tissue segmentation extremely difficult.

Methods

The comprehensive review of studies related to isointense brain MRI segmentation approaches is highlighted in this publication. The main aim and contribution of this study is to aid researchers by providing a thorough review to make their search for isointense brain MRI segmentation easier. The systematic literature review is performed from four points of reference: (1) review of studies concerning isointense brain MRI segmentation; (2) research contribution and future works and limitations; (3) frequently applied evaluation metrics and datasets; (4) findings of this studies.

Results and discussion

The systemic review is performed on studies that were published in the period of 2012 to 2022. A total of 19 primary studies of isointense brain MRI segmentation were selected to report the research question stated in this review.

Fully automated measurement of intracranial CSF and brain parenchyma volumes in pediatric hydrocephalus by segmentation of clinical MRI studies

BACKGROUND

Brain parenchyma (BP) and intracranial cerebrospinal fluid (iCSF) volumes measured by fully automated segmentation of clinical brain MRI studies may be useful for the diagnosis and follow-up of pediatric hydrocephalus. However, previously published segmentation techniques either rely on dedicated sequences, not routinely used in clinical practice, or on spatial normalization, which has limited accuracy when severe brain distortions, such as in hydrocephalic patients, are present.

PURPOSE

We developed a fully automated method to measure BP and iCSF volumes from clinical brain MRI studies of pediatric hydrocephalus patients, exploiting the complementary information contained in T2- and T1-weighted images commonly used in clinical practice.

METHODS

The proposed procedure, following skull-stripping of the combined volumes, performed using a multiparametric method to obtain a reliable definition of the inner skull profile, maximizes the CSF-to-parenchyma contrast by dividing the T2w- by the T1w- volume after full-scale dynamic rescaling, thus allowing separation of iCSF and BP through a simple thresholding routine.

RESULTS

Validation against manual tracing on 23 studies (four controls and 19 hydrocephalic patients) showed excellent concordance (ICC > 0.98) and spatial overlap (Dice coefficients ranging from 77.2% for iCSF to 96.8% for intracranial volume). Accuracy was comparable to the intra-operator reproducibility of manual segmentation, as measured in 14 studies processed twice by the same experienced neuroradiologist. Results of the application of the algorithm to a dataset of 63 controls and 57 hydrocephalic patients (19 with parenchymal damage), measuring volumes' changes with normal development and in hydrocephalic patients, are also reported for demonstration purposes.

CONCLUSIONS

The proposed approach allows fully automated segmentation of BP and iCSF in clinical studies, also in severely distorted brains, enabling to assess age- and disease-related changes in intracranial tissue volume with an accuracy comparable to expert manual segmentation.

The past, current, and future of neonatal intensive care units with artificial intelligence: a systematic review

Machine learning and deep learning are two subsets of artificial intelligence that involve teaching computers to learn and make decisions from any sort of data. Most recent developments in artificial intelligence are coming from deep learning, which has proven revolutionary in almost all fields, from computer vision to health sciences. The effects of deep learning in medicine have changed the conventional ways of clinical application significantly. Although some sub-fields of medicine, such as pediatrics, have been relatively slow in receiving the critical benefits of deep learning, related research in pediatrics has started to accumulate to a significant level, too. Hence, in this paper, we review recently developed machine learning and deep learning-based solutions for neonatology applications. We systematically evaluate the roles of both classical machine learning and deep learning in neonatology applications, define the methodologies, including algorithmic developments, and describe the remaining challenges in the assessment of neonatal diseases by using PRISMA 2020 guidelines. To date, the primary areas of focus in neonatology regarding AI applications have included survival analysis, neuroimaging, analysis of vital parameters and biosignals, and retinopathy of prematurity diagnosis. We have categorically summarized 106 research articles from 1996 to 2022 and discussed their pros and cons, respectively. In this systematic review, we aimed to further enhance the comprehensiveness of the study. We also discuss possible directions for new AI models and the future of neonatology with the rising power of AI, suggesting roadmaps for the integration of AI into neonatal intensive care units.

Early Childhood Development of Node Centrality in the White Matter Connectome and Its Relationship to IQ at Age 6 Years

BACKGROUND

The white matter (WM) connectome is important for cognitive development and intelligence and is altered in neuropsychiatric illnesses. Little is known about how the WM connectome develops or its relationship to IQ in early childhood.

METHODS

The development of node centrality in the WM connectome was studied in a longitudinal cohort of 226 (123 female) children from the University of North Carolina Early Brain Development Study. Structural and diffusion-weighted images were acquired after birth and at 1, 2, 4, and 6 years, and IQ was assessed at 6 years. Eigenvector centrality, betweenness centrality, and the global graph metrics of global efficiency, small worldness, and modularity were determined at each age.

RESULTS

The greatest developmental change in eigenvector centrality and betweenness centrality occurred during the first year of life, with relative stability between ages 1 and 6 years. Most of the high-centrality hubs at age 6 were also high-centrality hubs at 1 year, and many were already high-centrality hubs at birth. There were generally small but significant changes in global efficiency and modularity from birth to 6 years, while small worldness increased between 2 and 4 years. Individual node centrality was not significantly correlated with IQ at 6 years.

CONCLUSIONS

Node centrality in the WM connectome is established very early in childhood and is relatively stable from age 1 to 6 years. Many high-centrality hubs are established before birth, and most are present by age 1.

Review on deep learning fetal brain segmentation from Magnetic Resonance images

Brain segmentation is often the first and most critical step in quantitative analysis of the brain for many clinical applications, including fetal imaging. Different aspects challenge the segmentation of the fetal brain in magnetic resonance imaging (MRI), such as the non-standard position of the fetus owing to his/her movements during the examination, rapid brain development, and the limited availability of imaging data. In recent years, several segmentation methods have been proposed for automatically partitioning the fetal brain from MR images. These algorithms aim to define regions of interest with different shapes and intensities, encompassing the entire brain, or isolating specific structures. Deep learning techniques, particularly convolutional neural networks (CNNs), have become a state-of-the-art approach in the field because they can provide reliable segmentation results over heterogeneous datasets. Here, we review the deep learning algorithms developed in the field of fetal brain segmentation and categorize them according to their target structures. Finally, we discuss the perceived research gaps in the literature of the fetal domain, suggesting possible future research directions that could impact the management of fetal MR images.

Assessing the effects of head modelling errors and measurement noise on EEG source localization accuracy in preterm newborns: A single-subject study

The accuracy of electroencephalogram (EEG) source localization is compromised because of head modelling errors. In this study, we investigated the effect of inaccuracy in the conductivity of head tissues and head model structural deficiencies on the accuracy of EEG source analysis in premature neonates. A series of EEG forward and inverse simulations was performed by introducing structural deficiencies into the reference head models to generate test models, which were then used to investigate head modelling errors caused by cerebrospinal fluid (CSF) exclusion, lack of grey matter (GM)-white matter (WM) distinction, fontanel exclusion and inaccuracy in skull conductivity. The modelling errors were computed between forward and inverse solutions obtained using the reference and test models generated for each deficiency. Our results showed that the exclusion of CSF from the head model had a strong widespread effect on the accuracy of the EEG source localization with position errors lower than 4.17 mm. The GM and WM distinction also caused strong localization errors (up to 3.5 mm). The exclusion of fontanels from the head model also strongly affected the accuracy of the EEG source localization for sources located beneath the fontanels with a maximum localization error of 4.37 mm. Similarly, inaccuracies in the skull conductivity caused errors in EEG forward and inverse modelling in sources beneath cranial bones. Our results indicate that the accuracy of EEG source imaging in premature neonates can be largely improved by using head models, which include not only the brain, skull and scalp but also the CSF, GM, WM and fontanels.

Development of cortical folds in the human brain: An attempt to review biological hypotheses, early neuroimaging investigations and functional correlates

The folding of the human brain mostly takes place in utero, making it challenging to study. After a few pioneer studies looking into it in post-mortem foetal specimen, modern approaches based on neuroimaging have allowed the community to investigate the folding process in vivo, its normal progression, its early disturbances, and its relationship to later functional outcomes. In this review article, we aimed to first give an overview of the current hypotheses on the mechanisms governing cortical folding. After describing the methodological difficulties raised by its study in fetuses, neonates and infants with magnetic resonance imaging (MRI), we reported our current understanding of sulcal pattern emergence in the developing brain. We then highlighted the functional relevance of early sulcal development, through recent insights about hemispheric asymmetries and early factors influencing this dynamic such as prematurity. Finally, we outlined how longitudinal studies have started to relate early folding markers and the child's sensorimotor and cognitive outcome. Through this review, we hope to raise awareness on the potential of studying early sulcal patterns both from a fundamental and clinical perspective, as a window into early neurodevelopment and plasticity in relation to growth in utero and postnatal environment of the child.

An efficient memory reserving-and-fading strategy for vector quantization based 3D brain segmentation and tumor extraction using an unsupervised deep learning network

Deep learning networks are state-of-the-art approaches for 3D brain image segmentation, and the radiological characteristics extracted from tumors are of great significance for clinical diagnosis, treatment planning, and treatment outcome evaluation. However, two problems have been the hindering factors in brain image segmentation techniques. One is that deep learning networks require large amounts of manually annotated data. Another issue is the computational efficiency of 3D deep learning networks. In this study, we propose a vector quantization (VQ)-based 3D segmentation method that employs a novel unsupervised 3D deep embedding clustering (3D-DEC) network and an efficiency memory reserving-and-fading strategy. The VQ-based 3D-DEC network is trained on volume data in an unsupervised manner to avoid manual data annotation. The memory reserving-and-fading strategy beefs up model efficiency greatly. The designed methodology makes deep learning-based model feasible for biomedical image segmentation. The experiment is divided into two parts. First, we extensively evaluate the effectiveness and robustness of the proposed model on two authoritative MRI brain tumor databases (i.e., IBSR and BrainWeb). Second, we validate the model using real 3D brain tumor data collected from our institute for clinical practice significance. Results show that our method (without data manual annotation) has superior accuracy (0.74 ± 0.04 Tanimoto coefficient on IBSR, 97.5% TP and 97.7% TN on BrainWeb, and 91% Dice, 88% sensitivity and 87% specificity on real brain data) and remarkable efficiency (speedup ratio is 18.72 on IBSR, 31.16 on BrainWeb, 31.00 on real brain data) compared to the state-of-the-art methods. The results show that our proposed model can address the lacks of manual annotations, and greatly increase computation speedup with competitive segmentation accuracy compared to other state-of-the-art 3D CNN models. Moreover, the proposed model can be used for tumor treatment follow-ups every 6 months, providing critical details for surgical and postoperative treatment by correctly extracting numerical radiomic features of tumors.

Medical matting: Medical image segmentation with uncertainty from the matting perspective

High-quality manual labeling of ambiguous and complex-shaped targets with binary masks can be challenging. The weakness of insufficient expression of binary masks is prominent in segmentation, especially in medical scenarios where blurring is prevalent. Thus, reaching a consensus among clinicians through binary masks is more difficult in multi-person labeling cases. These inconsistent or uncertain areas are related to the lesions' structure and may contain anatomical information conducive to providing an accurate diagnosis. However, recent research focuses on uncertainties of model training and data labeling. None of them has investigated the influence of the ambiguous nature of the lesion itself. Inspired by image matting, this paper introduces a soft mask called alpha matte to medical scenes. It can describe the lesions with more details better than a binary mask. Moreover, it can also be used as a new uncertainty quantification method to represent uncertain areas, filling the gap in research on the uncertainty of lesion structure. In this work, we introduce a multi-task framework to generate binary masks and alpha mattes, which outperforms all state-of-the-art matting algorithms compared. The uncertainty map is proposed to imitate the trimap in matting methods, which can highlight fuzzy areas and improve matting performance. We have created three medical datasets with alpha mattes to address the lack of available matting datasets in medical fields and evaluated the effectiveness of our proposed method on them comprehensively. Furthermore, experiments demonstrate that the alpha matte is a more effective labeling method than the binary mask from both qualitative and quantitative aspects.

Examining time-frequency mechanisms of full-fledged deep sleep development in newborns of different gestational age in the first days of their postnatal development

Early age-related changes in EEG time-frequency characteristics during the restful sleep of newborns of different gestational ages result in the development of conventional EEG signs of deep sleep already during the first postnatal week of their life. Allocating newborns to different groups based on their gestational age and duration of postnatal period allowed demonstrating substantial intergroup differences in brain activity during sleep and wakefulness, along with significant variability in the time-frequency characteristics of brain activity. The process of conventional deep sleep development in infants born prior to the week 35 of gestation is associated with an increase in the power of alpha activity in the sensorimotor cortex of the brain.

PTNet3D: A 3D High-Resolution Longitudinal Infant Brain MRI Synthesizer Based on Transformers

An increased interest in longitudinal neurodevelopment during the first few years after birth has emerged in recent years. Noninvasive magnetic resonance imaging (MRI) can provide crucial information about the development of brain structures in the early months of life. Despite the success of MRI collections and analysis for adults, it remains a challenge for researchers to collect high-quality multimodal MRIs from developing infant brains because of their irregular sleep pattern, limited attention, inability to follow instructions to stay still during scanning. In addition, there are limited analytic approaches available. These challenges often lead to a significant reduction of usable MRI scans and pose a problem for modeling neurodevelopmental trajectories. Researchers have explored solving this problem by synthesizing realistic MRIs to replace corrupted ones. Among synthesis methods, the convolutional neural network-based (CNN-based) generative adversarial networks (GANs) have demonstrated promising performance. In this study, we introduced a novel 3D MRI synthesis framework- pyramid transformer network (PTNet3D)- which relies on attention mechanisms through transformer and performer layers. We conducted extensive experiments on high-resolution Developing Human Connectome Project (dHCP) and longitudinal Baby Connectome Project (BCP) datasets. Compared with CNN-based GANs, PTNet3D consistently shows superior synthesis accuracy and superior generalization on two independent, large-scale infant brain MRI datasets. Notably, we demonstrate that PTNet3D synthesized more realistic scans than CNN-based models when the input is from multi-age subjects. Potential applications of PTNet3D include synthesizing corrupted or missing images. By replacing corrupted scans with synthesized ones, we observed significant improvement in infant whole brain segmentation.

Newborn differential DNA methylation and subcortical brain volumes as early signs of severe neurodevelopmental delay in a South African Birth Cohort Study

OBJECTIVES

Early detection of neurodevelopmental delay is crucial for intervention and treatment strategies. We analysed associations between newborn DNA methylation (DNAm), neonatal magnetic resonance imaging (MRI) neuroimaging data, and neurodevelopment.

METHODS

Neurodevelopment was assessed in 161 children from the South African Drakenstein Child Health Study at 2 years of age using the Bayley Scales of Infant and Toddler Development III. We performed an epigenome-wide association study of neurodevelopmental delay using DNAm from cord blood. Subsequently, we analysed if associations between DNAm and neurodevelopmental delay were mediated by altered neonatal brain volumes (subset of 51 children).

RESULTS

Differential DNAm at SPTBN4 (cg26971411, Δbeta = -0.024, p-value = 3.28 × 10-08), and two intergenic regions (chromosome 11: cg00490349, Δbeta = -0.036, p-value = 3.02 × 10-08; chromosome 17: cg15660740, Δbeta = -0.078, p-value = 6.49 × 10-08) were significantly associated with severe neurodevelopmental delay. While these associations were not mediated by neonatal brain volume, neonatal caudate volumes were independently associated with neurodevelopmental delay, particularly in language (Δcaudate volume = 165.30 mm3, p = 0.0443) and motor (Δcaudate volume = 365.36 mm3, p-value = 0.0082) domains.

CONCLUSIONS

Differential DNAm from cord blood and increased neonatal caudate volumes were independently associated with severe neurodevelopmental delay at 2 years of age. These findings suggest that neurobiological signals for severe developmental delay may be detectable in very early life.

Cortical Source Analysis of the Face Sensitive N290 ERP Component in Infants at High Risk for Autism

Appropriate head models for cortical source analysis were investigated and applied to source analyses examining the neural bases of the face-sensitive N290 event-related potential (ERP) component in infants at high risk for autism spectrum disorder (ASD). This included infant siblings of children with ASD (ASIBs) and infants with fragile X syndrome (FXS). First, alternative head models for use with ASIBs and FXS were investigated. Head models created from the infant's own MRI were examined in relation to five head models based on average MRI templates. The results of the head model comparison identified group-specific (i.e., ASIB or FXS) head models created from a large collection of structural MRIs as the best substitution for the head model created from the participant's own structural MRI. Second, the cortical source analysis was completed on N290 data collected from a previous study to investigate brain areas associated with face sensitive ERP responses. Participants' own MRIs were used for head models when available, and the group-specific head model was used when the participants' own MRIs were not available. The results provide evidence for unique patterns of neural activation during face processing across infants at high and low risk for ASD and across etiologically distinct high-risk groups. All infants demonstrated greater activation to faces than toys in brain areas most associated with specialized face processing. Infants with FXS displayed higher levels of activation to faces across all areas analyzed, while ASIBs show more muted levels of activation. Overall, the results of the current study demonstrate the importance of group-specific head models for accurate cortical source analysis in infants at high risk for ASD. This also allows for further research on early distinctions in brain function based on risk status.

ID-Seg: an infant deep learning-based segmentation framework to improve limbic structure estimates

Infant brain magnetic resonance imaging (MRI) is a promising approach for studying early neurodevelopment. However, segmenting small regions such as limbic structures is challenging due to their low inter-regional contrast and high curvature. MRI studies of the adult brain have successfully applied deep learning techniques to segment limbic structures, and similar deep learning models are being leveraged for infant studies. However, these deep learning-based infant MRI segmentation models have generally been derived from small datasets, and may suffer from generalization problems. Moreover, the accuracy of segmentations derived from these deep learning models relative to more standard Expectation-Maximization approaches has not been characterized. To address these challenges, we leveraged a large, public infant MRI dataset (n = 473) and the transfer-learning technique to first pre-train a deep convolutional neural network model on two limbic structures: amygdala and hippocampus. Then we used a leave-one-out cross-validation strategy to fine-tune the pre-trained model and evaluated it separately on two independent datasets with manual labels. We term this new approach the Infant Deep learning SEGmentation Framework (ID-Seg). ID-Seg performed well on both datasets with a mean dice similarity score (DSC) of 0.87, a mean intra-class correlation (ICC) of 0.93, and a mean average surface distance (ASD) of 0.31 mm. Compared to the Developmental Human Connectome pipeline (dHCP) pipeline, ID-Seg significantly improved segmentation accuracy. In a third infant MRI dataset (n = 50), we used ID-Seg and dHCP separately to estimate amygdala and hippocampus volumes and shapes. The estimates derived from ID-seg, relative to those from the dHCP, showed stronger associations with behavioral problems assessed in these infants at age 2. In sum, ID-Seg consistently performed well on two different datasets with an 0.87 DSC, however, multi-site testing and extension for brain regions beyond the amygdala and hippocampus are still needed.

Segmentation of Infant Brain Using Nonnegative Matrix Factorization

This study develops an atlas-based automated framework for segmenting infants’ brains from magnetic resonance imaging (MRI). For the accurate segmentation of different structures of an infant’s brain at the isointense age (6–12 months), our framework integrates features of diffusion tensor imaging (DTI) (e.g., the fractional anisotropy (FA)). A brain diffusion tensor (DT) image and its region map are considered samples of a Markov–Gibbs random field (MGRF) that jointly models visual appearance, shape, and spatial homogeneity of a goal structure. The visual appearance is modeled with an empirical distribution of the probability of the DTI features, fused by their nonnegative matrix factorization (NMF) and allocation to data clusters. Projecting an initial high-dimensional feature space onto a low-dimensional space of the significant fused features with the NMF allows for better separation of the goal structure and its background. The cluster centers in the latter space are determined at the training stage by the K-means clustering. In order to adapt to large infant brain inhomogeneities and segment the brain images more accurately, appearance descriptors of both the first-order and second-order are taken into account in the fused NMF feature space. Additionally, a second-order MGRF model is used to describe the appearance based on the voxel intensities and their pairwise spatial dependencies. An adaptive shape prior that is spatially variant is constructed from a training set of co-aligned images, forming an atlas database. Moreover, the spatial homogeneity of the shape is described with a spatially uniform 3D MGRF of the second-order for region labels. In vivo experiments on nine infant datasets showed promising results in terms of the accuracy, which was computed using three metrics: the 95-percentile modified Hausdorff distance (MHD), the Dice similarity coefficient (DSC), and the absolute volume difference (AVD). Both the quantitative and visual assessments confirm that integrating the proposed NMF-fused DTI feature and intensity MGRF models of visual appearance, the adaptive shape prior, and the shape homogeneity MGRF model is promising in segmenting the infant brain DTI.

Recurrent Tissue-Aware Network for Deformable Registration of Infant Brain MR Images

Deformable registration is fundamental to longitudinal and population-based image analyses. However, it is challenging to precisely align longitudinal infant brain MR images of the same subject, as well as cross-sectional infant brain MR images of different subjects, due to fast brain development during infancy. In this paper, we propose a recurrently usable deep neural network for the registration of infant brain MR images. There are three main highlights of our proposed method. (i) We use brain tissue segmentation maps for registration, instead of intensity images, to tackle the issue of rapid contrast changes of brain tissues during the first year of life. (ii) A single registration network is trained in a one-shot manner, and then recurrently applied in inference for multiple times, such that the complex deformation field can be recovered incrementally. (iii) We also propose both the adaptive smoothing layer and the tissue-aware anti-folding constraint into the registration network to ensure the physiological plausibility of estimated deformations without degrading the registration accuracy. Experimental results, in comparison to the state-of-the-art registration methods, indicate that our proposed method achieves the highest registration accuracy while still preserving the smoothness of the deformation field. The implementation of our proposed registration network is available online https://github.com/Barnonewdm/ACTA-Reg-Net.

A 4D infant brain volumetric atlas based on the UNC/UMN baby connectome project (BCP) cohort

Spatiotemporal (four-dimensional) infant-dedicated brain atlases are essential for neuroimaging analysis of early dynamic brain development. However, due to the substantial technical challenges in the acquisition and processing of infant brain MR images, 4D atlases densely covering the dynamic brain development during infancy are still scarce. Few existing ones generally have fuzzy tissue contrast and low spatiotemporal resolution, leading to degraded accuracy of atlas-based normalization and subsequent analyses. To address this issue, in this paper, we construct a 4D structural MRI atlas for infant brains based on the UNC/UMN Baby Connectome Project (BCP) dataset, which features a high spatial resolution, extensive age-range coverage, and densely sampled time points. Specifically, 542 longitudinal T1w and T2w scans from 240 typically developing infants up to 26-month of age were utilized for our atlas construction. To improve the co-registration accuracy of the infant brain images, which typically exhibit dynamic appearance with low tissue contrast, we employed the state-of-the-art registration method and leveraged our generated reliable brain tissue probability maps in addition to the intensity images to improve the alignment of individual images. To achieve consistent region labeling on both infant and adult brain images for facilitating region-based analysis across ages, we mapped the widely used Desikan cortical parcellation onto our atlas by following an age-decreasing mapping manner. Meanwhile, the typical subcortical structures were manually delineated to facilitate the studies related to the subcortex. Compared with the existing infant brain atlases, our 4D atlas has much higher spatiotemporal resolution and preserves more structural details, and thus can boost accuracy in neurodevelopmental analysis during infancy.

Dear reviewers: Responses to common reviewer critiques about infant neuroimaging studies

The field of adult neuroimaging relies on well-established principles in research design, imaging sequences, processing pipelines, as well as safety and data collection protocols. The field of infant magnetic resonance imaging, by comparison, is a young field with tremendous scientific potential but continuously evolving standards. The present article aims to initiate a constructive dialog between researchers who grapple with the challenges and inherent limitations of a nascent field and reviewers who evaluate their work. We address 20 questions that researchers commonly receive from research ethics boards, grant, and manuscript reviewers related to infant neuroimaging data collection, safety protocols, study planning, imaging sequences, decisions related to software and hardware, and data processing and sharing, while acknowledging both the accomplishments of the field and areas of much needed future advancements. This article reflects the cumulative knowledge of experts in the FIT'NG community and can act as a resource for both researchers and reviewers alike seeking a deeper understanding of the standards and tradeoffs involved in infant neuroimaging.

Placental genomic risk scores and early neurodevelopmental outcomes

Tracing the early paths leading to developmental disorders is critical for prevention. In previous work, we detected an interaction between genomic risk scores for schizophrenia (GRSs) and early-life complications (ELCs), so that the liability of the disorder explained by genomic risk was higher in the presence of a history of ELCs, compared with its absence. This interaction was specifically driven by loci harboring genes highly expressed in placentae from normal and complicated pregnancies [G. Ursini et al., Nat. Med. 24, 792-801 (2018)]. Here, we analyze whether fractionated genomic risk scores for schizophrenia and other developmental disorders and traits, based on placental gene-expression loci (PlacGRSs), are linked with early neurodevelopmental outcomes in individuals with a history of ELCs. We found that schizophrenia's PlacGRSs are negatively associated with neonatal brain volume in singletons and offspring of multiple pregnancies and, in singletons, with cognitive development at 1 y and, less strongly, at 2 y, when cognitive scores become more sensitive to other factors. These negative associations are stronger in males, found only with GRSs fractionated by placental gene expression, and not found in PlacGRSs for other developmental disorders and traits. The relationship of PlacGRSs with brain volume persists as an anlage of placenta biology in adults with schizophrenia, again selectively in males. Higher placental genomic risk for schizophrenia, in the presence of ELCs and particularly in males, alters early brain growth and function, defining a potentially reversible neurodevelopmental path of risk that may be unique to schizophrenia.

Morphometric Analysis of Brain in Newborn with Congenital Diaphragmatic Hernia

Congenital diaphragmatic hernia (CDH) is a severe pediatric disorder with herniation of abdominal viscera into the thoracic cavity. Since neurodevelopmental impairment constitutes a common outcome, we performed morphometric magnetic resonance imaging (MRI) analysis on CDH infants to investigate cortical parameters such as cortical thickness (CT) and local gyrification index (LGI). By assessing CT and LGI distributions and their correlations with variables which might have an impact on oxygen delivery (total lung volume, TLV), we aimed to detect how altered perfusion affects cortical development in CDH. A group of CDH patients received both prenatal (i.e., fetal stage) and postnatal MRI. From postnatal high-resolution T2-weighted images, mean CT and LGI distributions of 16 CDH were computed and statistically compared to those of 13 controls. Moreover, TLV measures obtained from fetal MRI were further correlated to LGI. Compared to controls, CDH infants exhibited areas of hypogiria within bilateral fronto-temporo-parietal labels, while no differences were found for CT. LGI significantly correlated with TLV within bilateral temporal lobes and left frontal lobe, involving language- and auditory-related brain areas. Although the causes of neurodevelopmental impairment in CDH are still unclear, our results may suggest their link with altered cortical maturation and possible impaired oxygen perfusion.

Imaging the rapidly developing brain: Current challenges for MRI studies in the first five years of life

Rapid and widespread changes in brain anatomy and physiology in the first five years of life present substantial challenges for developmental structural, functional, and diffusion MRI studies. One persistent challenge is that methods best suited to earlier developmental stages are suboptimal for later stages, which engenders a trade-off between using different, but age-appropriate, methods for different developmental stages or identical methods across stages. Both options have potential benefits, but also biases, as pipelines for each developmental stage can be matched on methods or the age-appropriateness of methods, but not both. This review describes the data acquisition, processing, and analysis challenges that introduce these potential biases and attempts to elucidate decisions and make recommendations that would optimize developmental comparisons.

Intuitionistic Fuzzy C-Means Algorithm Based on Membership Information Transfer-Ring and Similarity Measurement

The fuzzy C-means clustering (FCM) algorithm is used widely in medical image segmentation and suitable for segmenting brain tumors. Therefore, an intuitionistic fuzzy C-means algorithm based on membership information transferring and similarity measurements (IFCM-MS) is proposed to segment brain tumor magnetic resonance images (MRI) in this paper. The original FCM lacks spatial information, which leads to a high noise sensitivity. To address this issue, the membership information transfer model is adopted to the IFCM-MS. Specifically, neighborhood information and the similarity of adjacent iterations are incorporated into the clustering process. Besides, FCM uses simple distance measurements to calculate the membership degree, which causes an unsatisfactory result. So, a similarity measurement method is designed in the IFCM-MS to improve the membership calculation, in which gray information and distance information are fused adaptively. In addition, the complex structure of the brain results in MRIs with uncertainty boundary tissues. To overcome this problem, an intuitive fuzzy attribute is embedded into the IFCM-MS. Experiments performed on real brain tumor images demonstrate that our IFCM-MS has low noise sensitivity and high segmentation accuracy.

Development, validation, and pilot MRI safety study of a high-resolution, open source, whole body pediatric numerical simulation model

Numerical body models of children are used for designing medical devices, including but not limited to optical imaging, ultrasound, CT, EEG/MEG, and MRI. These models are used in many clinical and neuroscience research applications, such as radiation safety dosimetric studies and source localization. Although several such adult models have been reported, there are few reports of full-body pediatric models, and those described have several limitations. Some, for example, are either morphed from older children or do not have detailed segmentations. Here, we introduce a 29-month-old male whole-body native numerical model, "MARTIN", that includes 28 head and 86 body tissue compartments, segmented directly from the high spatial resolution MRI and CT images. An advanced auto-segmentation tool was used for the deep-brain structures, whereas 3D Slicer was used to segment the non-brain structures and to refine the segmentation for all of the tissue compartments. Our MARTIN model was developed and validated using three separate approaches, through an iterative process, as follows. First, the calculated volumes, weights, and dimensions of selected structures were adjusted and confirmed to be within 6% of the literature values for the 2-3-year-old age-range. Second, all structural segmentations were adjusted and confirmed by two experienced, sub-specialty certified neuro-radiologists, also through an interactive process. Third, an additional validation was performed with a Bloch simulator to create synthetic MR image from our MARTIN model and compare the image contrast of the resulting synthetic image with that of the original MRI data; this resulted in a "structural resemblance" index of 0.97. Finally, we used our model to perform pilot MRI safety simulations of an Active Implantable Medical Device (AIMD) using a commercially available software platform (Sim4Life), incorporating the latest International Standards Organization guidelines. This model will be made available on the Athinoula A. Martinos Center for Biomedical Imaging website.

Effect of structural complexities in head modeling on the accuracy of EEG source localization in neonates

OBJECTIVE

Neonatal electroencephalography (EEG) source localization is highly prone to errors due to head modeling deficiencies. In this study, we investigated the effect of head model complexities on the accuracy of EEG source localization in full term neonates using a realistic volume conductor head model.

APPROACH

We performed numerical simulations to investigate source localization errors caused by cerebrospinal fluid (CSF) and fontanel exclusion and gray matter (GM)/white matter (WM) distinction using the finite element method.

MAIN RESULTS

Our results showed that the exclusion of CSF from the head model could cause significant localization errors mostly for sources closer to the inner surface of the skull. With a less pronounced effect compared to the CSF exclusion, the discrimination between GM and WM also widely affected all sources, especially those located in deeper structures. The exclusion of the fontanels from the head model led to source localization errors for sources located in areas beneath the fontanels. Our finding clearly shows that the CSF inclusion and GM/WM distinction in EEG inverse modeling can substantially reduce EEG source localization errors. Moreover, fontanels should be included in neonatal head models, particularly in source localization applications, in which sources of interest are located beneath or in vicinity of fontanels.

SIGNIFICANCE

Our findings have practical implications for a better understanding of the impact of head model complexities on the accuracy of EEG source localization in neonates.

Neonatal brain volume as a marker of differential susceptibility to parenting quality and its association with neurodevelopment across early childhood

Parenting quality is associated with child cognitive and executive functions (EF), which are important predictors of social and academic development. However, children vary in their susceptibility to parenting behaviors, and the neurobiological underpinnings of this susceptibility are poorly understood. In a prospective longitudinal study, we examined whether neonatal total brain volume (TBV) and subregions of interest (i.e., hippocampus (HC) and anterior cingulate gyrus (ACG)) moderate the association between maternal sensitivity and cognitive/EF development across early childhood. Neonates underwent a brain magnetic resonance imaging scan. Their cognitive performance and EF was characterized at 2.0 ± 0.1 years (N = 53) and at 4.9 ± 0.8 years (N = 36) of age. Maternal sensitivity was coded based on observation of a standardized play situation at 6-mo postpartum. Neonatal TBV moderated the association between maternal sensitivity and 2-year working memory as well as all 5-year cognitive outcomes, suggesting that the positive association between maternal sensitivity and child cognition was observed only among children with large or average but not small TBV as neonates. Similar patterns were observed for TBV-corrected HC and ACG volumes. The findings suggest that larger neonatal TBV, HC and ACG may underlie susceptibility to the environment and affect the degree to which parenting quality shapes long-term cognitive development.

Individual Variation of Human Cortical Structure Is Established in the First Year of Life

BACKGROUND

Individual differences in cortical gray matter (GM) structure are associated with cognitive function and psychiatric disorders with developmental origins. Identifying when individual differences in cortical structure are established in childhood is critical for understanding the timing of abnormal cortical development associated with neuropsychiatric disorders.

METHODS

We studied the development of cortical GM and white matter volume, cortical thickness, and surface area using structural magnetic resonance imaging in two unique cohorts of singleton (121 male and 131 female) and twin (99 male and 83 female) children imaged longitudinally from birth to 6 years.

RESULTS

Cortical GM volume increases rapidly in the first year of life, with more gradual growth thereafter. Between ages 1 and 6 years, total surface area expands 29%, while average cortical thickness decreases about 3.5%. In both cohorts, a large portion of individual variation in cortical GM volume (81%-87%) and total surface area (73%-83%) at age 6 years is present by age 1 year. Regional heterogeneity of cortical thickness observed at age 6 is largely in place at age 1.

CONCLUSIONS

These findings indicate that individual differences in cortical GM structure are largely established by the end of the first year of life, following a period of rapid postnatal GM growth. This suggests that alterations in GM structure associated with psychiatric disorders with developmental origins may largely arise in the first year of life and that interventions to normalize or mitigate abnormal GM development may need to be targeted to very early childhood.

Infant FreeSurfer: An automated segmentation and surface extraction pipeline for T1-weighted neuroimaging data of infants 0-2 years

The development of automated tools for brain morphometric analysis in infants has lagged significantly behind analogous tools for adults. This gap reflects the greater challenges in this domain due to: 1) a smaller-scaled region of interest, 2) increased motion corruption, 3) regional changes in geometry due to heterochronous growth, and 4) regional variations in contrast properties corresponding to ongoing myelination and other maturation processes. Nevertheless, there is a great need for automated image-processing tools to quantify differences between infant groups and other individuals, because aberrant cortical morphologic measurements (including volume, thickness, surface area, and curvature) have been associated with neuropsychiatric, neurologic, and developmental disorders in children. In this paper we present an automated segmentation and surface extraction pipeline designed to accommodate clinical MRI studies of infant brains in a population 0-2 year-olds. The algorithm relies on a single channel of T1-weighted MR images to achieve automated segmentation of cortical and subcortical brain areas, producing volumes of subcortical structures and surface models of the cerebral cortex. We evaluated the algorithm both qualitatively and quantitatively using manually labeled datasets, relevant comparator software solutions cited in the literature, and expert evaluations. The computational tools and atlases described in this paper will be distributed to the research community as part of the FreeSurfer image analysis package.

Rhesus Macaque Brain Developmental Trajectory: A Longitudinal Analysis Using Tensor-Based Structural Morphometry and Diffusion Tensor Imaging

The typical developmental trajectory of brain structure among nonhuman primates (NHPs) remains poorly understood. In this study, we characterized the normative trajectory of developmental change among a cohort of rhesus monkeys (n = 28), ranging in age from 2 to 22 months, using structural MRI datasets that were longitudinally acquired every 3-4 months. We hypothesized that NHP-specific transient intracranial volume decreases reported during late infancy would be part of the typical developmental process, which is driven by volumetric contraction of gray matter in primary functional areas. To this end, we performed multiscale analyses from the whole brain to voxel level, characterizing regional heterogeneity, hemispheric asymmetry, and sexual dimorphism in developmental patterns. The longitudinal trajectory of brain development was explained by three different regional volumetric growth patterns (exponentially decreasing, undulating, and linearly increasing), which resulted in developmental brain volume curves with transient brain volumetric decreases. White matter (WM) fractional anisotropy increased with age, corresponding to WM volume increases, while mean diffusivity (MD) showed biphasic patterns. The longitudinal trajectory of brain development in young rhesus monkeys follows typical maturation patterns seen in humans, but regional volumetric and MD changes are more dynamic in rhesus monkeys compared with humans, with marked decreases followed by "rebound-like" increases.

SegSRGAN: Super-resolution and segmentation using generative adversarial networks - Application to neonatal brain MRI

BACKGROUND AND OBJECTIVE

One of the main issues in the analysis of clinical neonatal brain MRI is the low anisotropic resolution of the data. In most MRI analysis pipelines, data are first re-sampled using interpolation or single image super-resolution techniques and then segmented using (semi-)automated approaches. In other words, image reconstruction and segmentation are then performed separately. In this article, we propose a methodology and a software solution for carrying out simultaneously high-resolution reconstruction and segmentation of brain MRI data.

METHODS

Our strategy mainly relies on generative adversarial networks. The network architecture is described in detail. We provide information about its implementation, focusing on the most crucial technical points (whereas complementary details are given in a dedicated GitHub repository). We illustrate the behavior of the proposed method for cortex analysis from neonatal MR images.

RESULTS

The results of the method, evaluated quantitatively (Dice, peak signal-to-noise ratio, structural similarity, number of connected components) and qualitatively on a research dataset (dHCP) and a clinical one (Epirmex), emphasize the relevance of the approach, and its ability to take advantage of data-augmentation strategies.

CONCLUSIONS

Results emphasize the potential of our proposed method/software with respect to practical medical applications. The method is provided as a freely available software tool, which allows one to carry out his/her own experiments, and involve the method for the super-resolution reconstruction and segmentation of arbitrary cerebral structures from any MR image dataset.

Using Deep Convolutional Neural Networks for Neonatal Brain Image Segmentation

INTRODUCTION

Deep learning neural networks are especially potent at dealing with structured data, such as images and volumes. Both modified LiviaNET and HyperDense-Net performed well at a prior competition segmenting 6-month-old infant magnetic resonance images, but neonatal cerebral tissue type identification is challenging given its uniquely inverted tissue contrasts. The current study aims to evaluate the two architectures to segment neonatal brain tissue types at term equivalent age.

METHODS

Both networks were retrained over 24 pairs of neonatal T1 and T2 data from the Developing Human Connectome Project public data set and validated on another eight pairs against ground truth. We then reported the best-performing model from training and its performance by computing the Dice similarity coefficient (DSC) for each tissue type against eight test subjects.

RESULTS

During the testing phase, among the segmentation approaches tested, the dual-modality HyperDense-Net achieved the best statistically significantly test mean DSC values, obtaining 0.94/0.95/0.92 for the tissue types and took 80 h to train and 10 min to segment, including preprocessing. The single-modality LiviaNET was better at processing T2-weighted images than processing T1-weighted images across all tissue types, achieving mean DSC values of 0.90/0.90/0.88 for gray matter, white matter, and cerebrospinal fluid, respectively, while requiring 30 h to train and 8 min to segment each brain, including preprocessing.

DISCUSSION

Our evaluation demonstrates that both neural networks can segment neonatal brains, achieving previously reported performance. Both networks will be continuously retrained over an increasingly larger repertoire of neonatal brain data and be made available through the Canadian Neonatal Brain Platform to better serve the neonatal brain imaging research community.

Automatic Measurement of Extra-Axial CSF from Infant MRI Data

The quantification of cerebrospinal fluid (CSF) in the human brain has shown to play an important role in early postnatal brain development. Extra-axial fluid (EA-CSF), which is characterized by CSF in the subarachnoid space, is a promising marker for the early detection of children at risk for neurodevelopmental disorders, such as Autism Spectrum Disorder (ASD). Yet, non-ventricular CSF quantification, in particular extra-axial CSF quantification, is not supported in the major neuro-imaging software solutions, such as FreeSurfer. Most current structural image analysis packages mask out the extra-axial CSF space in one of the first pre-processing steps. A quantitative protocol was previously developed by our group to objectively measure the volume of total EA-CSF volume using a pipeline workflow implemented in a series of python scripts. While this solution worked for our specific lab, a graphical user interface-based tool is necessary to facilitate the computation of extra-axial CSF volume across a wide array of neuroimaging studies and research labs. This paper presents the development of a novel open-source, cross-platform, user-friendly software tool, called Auto-EACSF, for the automatic computation of such extra-axial CSF volume. Auto-EACSF allows neuroimaging labs to quantify extra-axial CSF in their neuroimaging studies in order to investigate its role in normal and atypical brain development.

Maturational Changes in Human Dorsal and Ventral Visual Networks

Developmental neuroimaging studies report the emergence of increasingly diverse cognitive functions as closely entangled with a rise-fall modulation of cortical thickness (CTh), structural cortical and white-matter connectivity, and a time-course for the experience-dependent selective elimination of the overproduced synapses. We examine which of two visual processing networks, the dorsal (DVN; prefrontal, parietal nodes) or ventral (VVN; frontal-temporal, fusiform nodes) matures first, thus leading the neuro-cognitive developmental trajectory. Three age-dependent measures are reported: (i) the CTh at network nodes; (ii) the matrix of intra-network structural connectivity (edges); and (iii) the proficiency in network-related neuropsychological tests. Typically developing children (age ~6 years), adolescents (~11 years), and adults (~21 years) were tested using multiple-acquisition structural T1-weighted magnetic resonance imaging (MRI) and neuropsychology. MRI images reconstructed into a gray/white/pial matter boundary model were used for CTh evaluation. No significant group differences in CTh and in the matrix of edges were found for DVN (except for the left prefrontal), but a significantly thicker cortex in children for VVN with reduced prefrontal ventral-fusiform connectivity and with an abundance of connections in adolescents. The higher performance in children on tests related to DVN corroborates the age-dependent MRI structural connectivity findings. The current findings are consistent with an earlier maturational course of DVN.

Deep CNN ensembles and suggestive annotations for infant brain MRI segmentation

Precise 3D segmentation of infant brain tissues is an essential step towards comprehensive volumetric studies and quantitative analysis of early brain development. However, computing such segmentations is very challenging, especially for 6-month infant brain, due to the poor image quality, among other difficulties inherent to infant brain MRI, e.g., the isointense contrast between white and gray matter and the severe partial volume effect due to small brain sizes. This study investigates the problem with an ensemble of semi-dense fully convolutional neural networks (CNNs), which employs T1-weighted and T2-weighted MR images as input. We demonstrate that the ensemble agreement is highly correlated with the segmentation errors. Therefore, our method provides measures that can guide local user corrections. To the best of our knowledge, this work is the first ensemble of 3D CNNs for suggesting annotations within images. Our quasi-dense architecture allows the efficient propagation of gradients during training, while limiting the number of parameters, requiring one order of magnitude less parameters than popular medical image segmentation networks such as 3D U-Net (Çiçek, et al.). We also investigated the impact that early or late fusions of multiple image modalities might have on the performances of deep architectures. We report evaluations of our method on the public data of the MICCAI iSEG-2017 Challenge on 6-month infant brain MRI segmentation, and show very competitive results among 21 teams, ranking first or second in most metrics.

TRActs constrained by UnderLying INfant anatomy (TRACULInA): An automated probabilistic tractography tool with anatomical priors for use in the newborn brain

The ongoing myelination of white-matter fiber bundles plays a significant role in brain development. However, reliable and consistent identification of these bundles from infant brain MRIs is often challenging due to inherently low diffusion anisotropy, as well as motion and other artifacts. In this paper we introduce a new tool for automated probabilistic tractography specifically designed for newborn infants. Our tool incorporates prior information about the anatomical neighborhood of white-matter pathways from a training data set. In our experiments, we evaluate this tool on data from both full-term and prematurely born infants and demonstrate that it can reconstruct known white-matter tracts in both groups robustly, even in the presence of differences between the training set and study subjects. Additionally, we evaluate it on a publicly available large data set of healthy term infants (UNC Early Brain Development Program). This paves the way for performing a host of sophisticated analyses in newborns that we have previously implemented for the adult brain, such as pointwise analysis along tracts and longitudinal analysis, in both health and disease.

A Novel Approach for Manual Segmentation of the Amygdala and Hippocampus in Neonate MRI

The gross anatomy of the infant brain at term is fairly similar to that of the adult brain, but structures are immature, and the brain undergoes rapid growth during the first 2 years of life. Neonate magnetic resonance (MR) images have different contrasts compared to adult images, and automated segmentation of brain magnetic resonance imaging (MRI) can thus be considered challenging as less software options are available. Despite this, most anatomical regions are identifiable and thus amenable to manual segmentation. In the current study, we developed a protocol for segmenting the amygdala and hippocampus in T2-weighted neonatal MR images. The participants were 31 healthy infants between 2 and 5 weeks of age. Intra-rater reliability was measured in 12 randomly selected MR images, where 6 MR images were segmented at 1-month intervals between the delineations, and another 6 MR images at 6-month intervals. The protocol was also tested by two independent raters in 20 randomly selected T2-weighted images, and finally with T1 images. Intraclass correlation coefficient (ICC) and Dice similarity coefficient (DSC) for intra-rater, inter-rater, and T1 vs. T2 comparisons were computed. Moreover, manual segmentations were compared to automated segmentations performed by iBEAT toolbox in 10 T2-weighted MR images. The intra-rater reliability was high ICC ≥ 0.91, DSC ≥ 0.89, the inter-rater reliabilities were satisfactory ICC ≥ 0.90, DSC ≥ 0.75 for hippocampus and DSC ≥ 0.52 for amygdalae. Segmentations for T1 vs. T2-weighted images showed high consistency ICC ≥ 0.90, DSC ≥ 0.74. The manual and iBEAT segmentations showed no agreement, DSC ≥ 0.39. In conclusion, there is a clear need to improve and develop the procedures for automated segmentation of infant brain MR images.

Quantifying deep grey matter atrophy using automated segmentation approaches: A systematic review of structural MRI studies

The deep grey matter (DGM) nuclei of the brain play a crucial role in learning, behaviour, cognition, movement and memory. Although automated segmentation strategies can provide insight into the impact of multiple neurological conditions affecting these structures, such as Multiple Sclerosis (MS), Huntington's disease (HD), Alzheimer's disease (AD), Parkinson's disease (PD) and Cerebral Palsy (CP), there are a number of technical challenges limiting an accurate automated segmentation of the DGM. Namely, the insufficient contrast of T1 sequences to completely identify the boundaries of these structures, as well as the presence of iso-intense white matter lesions or extensive tissue loss caused by brain injury. Therefore in this systematic review, 269 eligible studies were analysed and compared to determine the optimal approaches for addressing these technical challenges. The automated approaches used among the reviewed studies fall into three broad categories, atlas-based approaches focusing on the accurate alignment of atlas priors, algorithmic approaches which utilise intensity information to a greater extent, and learning-based approaches that require an annotated training set. Studies that utilise freely available software packages such as FIRST, FreeSurfer and LesionTOADS were also eligible, and their performance compared. Overall, deep learning approaches achieved the best overall performance, however these strategies are currently hampered by the lack of large-scale annotated data. Improving model generalisability to new datasets could be achieved in future studies with data augmentation and transfer learning. Multi-atlas approaches provided the second-best performance overall, and may be utilised to construct a "silver standard" annotated training set for deep learning. To address the technical challenges, providing robustness to injury can be improved by using multiple channels, highly elastic diffeomorphic transformations such as LDDMM, and by following atlas-based approaches with an intensity driven refinement of the segmentation, which has been done with the Expectation Maximisation (EM) and level sets methods. Accounting for potential lesions should be achieved with a separate lesion segmentation approach, as in LesionTOADS. Finally, to address the issue of limited contrast, R2*, T2* and QSM sequences could be used to better highlight the DGM due to its higher iron content. Future studies could look to additionally acquire these sequences by retaining the phase information from standard structural scans, or alternatively acquiring these sequences for only a training set, allowing models to learn the "improved" segmentation from T1-sequences alone.

HyperDense-Net: A Hyper-Densely Connected CNN for Multi-Modal Image Segmentation

Recently, dense connections have attracted substantial attention in computer vision because they facilitate gradient flow and implicit deep supervision during training. Particularly, DenseNet that connects each layer to every other layer in a feed-forward fashion and has shown impressive performances in natural image classification tasks. We propose HyperDenseNet, a 3-D fully convolutional neural network that extends the definition of dense connectivity to multi-modal segmentation problems. Each imaging modality has a path, and dense connections occur not only between the pairs of layers within the same path but also between those across different paths. This contrasts with the existing multi-modal CNN approaches, in which modeling several modalities relies entirely on a single joint layer (or level of abstraction) for fusion, typically either at the input or at the output of the network. Therefore, the proposed network has total freedom to learn more complex combinations between the modalities, within and in-between all the levels of abstraction, which increases significantly the learning representation. We report extensive evaluations over two different and highly competitive multi-modal brain tissue segmentation challenges, iSEG 2017 and MRBrainS 2013, with the former focusing on six month infant data and the latter on adult images. HyperDenseNet yielded significant improvements over many state-of-the-art segmentation networks, ranking at the top on both benchmarks. We further provide a comprehensive experimental analysis of features re-use, which confirms the importance of hyper-dense connections in multi-modal representation learning. Our code is publicly available.

Prediction of adverse motor outcome for neonates with punctate white matter lesions by MRI images using radiomics strategy: protocol for a prospective cohort multicentre study

INTRODUCTION

Punctate white matter lesions (PWML) are prevalent white matter disease in preterm neonates, and may cause motor disorders and even cerebral palsy. However, precise individual-based diagnosis of lesions that result in an adverse motor outcome remains unclear, and an effective method is urgently needed to guide clinical diagnosis and treatment. Advanced radiomics for multiple modalities data can provide a possible look for biomarkers and determine prognosis quantitatively. The study aims to develop and validate a model for prediction of adverse motor outcomes at a corrected age (CA) of 24 months in neonates with PWML.

METHODS AND ANALYSIS

A prospective cohort multicentre study will be conducted in 11 Chinese hospitals. A total of 394 neonates with PWML confirmed by MRI will undergo a clinical assessment (modified Neonatal Behavioural Assessment Scale). At a CA of 18 months, the motor function will be assessed by Bayley Scales of Infant and Toddler Development-III (Bayley-III). Mild-to-severe motor impairments will be confirmed using the Bayley-III and Gross Motor Function Classification System at a CA of 24 months. During the data collection, the perinatal and clinical information will also be recorded. According to the radiomics strategy, the extracted imaging features and clinical information will be combined for exploratory analysis. After using multiple-modelling methodology, the accuracy, sensitivity and specificity will be computed. Internal and external validations will be used to evaluate the performance of the radiomics model.

ETHICS AND DISSEMINATION

This study has been approved by the institutional review board of The First Affiliated Hospital of Xi'an Jiaotong University (XJTU1AF2015LSK-172). All parents of eligible participants will be provided with a detailed explanation of the study and written consent will be obtained. The results of this study will be published in peer-reviewed journals and presented at local, national and international conferences.

TRIAL REGISTRATION NUMBER

NCT02637817; Pre-results.

3-D Fully Convolutional Networks for Multimodal Isointense Infant Brain Image Segmentation

Accurate segmentation of infant brain images into different regions of interest is one of the most important fundamental steps in studying early brain development. In the isointense phase (approximately 6-8 months of age), white matter and gray matter exhibit similar levels of intensities in magnetic resonance (MR) images, due to the ongoing myelination and maturation. This results in extremely low tissue contrast and thus makes tissue segmentation very challenging. Existing methods for tissue segmentation in this isointense phase usually employ patch-based sparse labeling on single modality. To address the challenge, we propose a novel 3-D multimodal fully convolutional network (FCN) architecture for segmentation of isointense phase brain MR images. Specifically, we extend the conventional FCN architectures from 2-D to 3-D, and, rather than directly using FCN, we intuitively integrate coarse (naturally high-resolution) and dense (highly semantic) feature maps to better model tiny tissue regions, in addition, we further propose a transformation module to better connect the aggregating layers; we also propose a fusion module to better serve the fusion of feature maps. We compare the performance of our approach with several baseline and state-of-the-art methods on two sets of isointense phase brain images. The comparison results show that our proposed 3-D multimodal FCN model outperforms all previous methods by a large margin in terms of segmentation accuracy. In addition, the proposed framework also achieves faster segmentation results compared to all other methods. Our experiments further demonstrate that: 1) carefully integrating coarse and dense feature maps can considerably improve the segmentation performance; 2) batch normalization can speed up the convergence of the networks, especially when hierarchical feature aggregations occur; and 3) integrating multimodal information can further boost the segmentation performance.

Age-specific gray and white matter DTI atlas for human brain at 33, 36 and 39 postmenstrual weeks

During the 3rd trimester, dramatic structural changes take place in the human brain, underlying the neural circuit formation. The survival rate of premature infants has increased significantly in recent years. The large morphological differences of the preterm brain at 33 or 36 postmenstrual weeks (PMW) from the brain at 40PMW (full term) make it necessary to establish age-specific atlases for preterm brains. In this study, with high quality (1.5 × 1.5 × 1.6 mm3 imaging resolution) diffusion tensor imaging (DTI) data obtained from 84 healthy preterm and term-born neonates, we established age-specific preterm and term-born brain templates and atlases at 33, 36 and 39PMW. Age-specific DTI templates include a single-subject template, a population-averaged template with linear transformation and a population-averaged template with nonlinear transformation. Each of the age-specific DTI atlases includes comprehensive labeling of 126 major gray matter (GM) and white matter (WM) structures, specifically 52 cerebral cortical structures, 40 cerebral WM structures, 22 brainstem and cerebellar structures and 12 subcortical GM structures. From 33 to 39 PMW, dramatic morphological changes of delineated individual neural structures such as ganglionic eminence and uncinate fasciculus were revealed. The evaluation based on measurements of Dice ratio and L1 error suggested reliable and reproducible automated labels from the age-matched atlases compared to labels from manual delineation. Applying these atlases to automatically and effectively delineate microstructural changes of major WM tracts during the 3rd trimester was demonstrated. The established age-specific DTI templates and atlases of 33, 36 and 39 PMW brains may be used for not only understanding normal functional and structural maturational processes but also detecting biomarkers of neural disorders in the preterm brains.

Neural histology and neurogenesis of the human fetal and infant brain

The human brain develops slowly and over a long period of time which lasts for almost three decades. This enables good spatio-temporal resolution of histogenetic and neurogenetic events as well as an appropriate and clinically relevant timing of these events. In order to successfully apply in vivo neuroimaging data, in analyzing both the normal brain development and the neurodevelopmental origin of major neurological and mental disorders, it is important to correlate these neuroimaging data with the existing data on morphogenetic, histogenetic and neurogenetic events. Furthermore, when performing such correlation, the genetic, genomic, and molecular biology data on phenotypic specification of developing brain regions, areas and neurons should also be included. In this review, we focus on early developmental periods (form 8 postconceptional weeks to the second postnatal year) and describe the microstructural organization and neural circuitry elements of the fetal and early postnatal human cerebrum.