Research Articles, Development/Plasticity/Repair

Development of the Fetal Brain Corticocortical Structural Network during the Second-to-Third Trimester Based on Diffusion MRI

Ruike Chen, Ruoke Zhao, Haotian Li, Xinyi Xu, Mingyang Li, Zhiyong Zhao, Cong Sun, Guangbin Wang and Dan Wu
Journal of Neuroscience 17 July 2024, 44 (29) e1567232024; https://doi.org/10.1523/JNEUROSCI.1567-23.2024

Abstract

During the second-to-third trimester, the neuronal pathways of the fetal brain experience rapid development, resulting in the complex architecture of the interwired network at birth. While diffusion MRI-based tractography has been employed to study the prenatal development of structural connectivity network (SCN) in preterm neonatal and postmortem fetal brains, the in utero development of SCN in the normal fetal brain remains largely unknown. In this study, we utilized in utero dMRI data from human fetuses of both sexes between 26 and 38 gestational weeks to investigate the developmental trajectories of the fetal brain SCN, focusing on intrahemispheric connections. Our analysis revealed significant increases in global efficiency, mean local efficiency, and clustering coefficient, along with significant decrease in shortest path length, while small-worldness persisted during the studied period, revealing balanced network integration and segregation. Widespread short-ranged connectivity strengthened significantly. The nodal strength developed in a posterior-to-anterior and medial-to-lateral order, reflecting a spatiotemporal gradient in cortical network connectivity development. Moreover, we observed distinct lateralization patterns in the fetal brain SCN. Globally, there was a leftward lateralization in network efficiency, clustering coefficient, and small-worldness. The regional lateralization patterns in most language, motor, and visual-related areas were consistent with prior knowledge, except for Wernicke's area, indicating lateralized brain wiring is an innate property of the human brain starting from the fetal period. Our findings provided a comprehensive view of the development of the fetal brain SCN and its lateralization, as a normative template that may be used to characterize atypical development.

Significance Statement

We studied the normal development of intrahemispheric corticocortical structural connectivity networks (SCNs) of the fetal brain from 26 to 38 gestational weeks using in utero diffusion MRI data. Graph theory-based analysis revealed significant enhancement in network efficiency and clustering, as well as persistent small-worldness with age, revealing balanced integration and segregation in the fetal brain SCN during the studied period, supported by regional developmental patterns. Leftward lateralization in network efficiency, clustering coefficient, and small-worldness was observed. Regional lateralization patterns in most language, motor, and visual-related areas were consistent with prior knowledge. We also summarized the challenges of investigating the fetal brain SCN development and provided suggestions for future studies.

Introduction

The fetal brain neuronal pathways rapidly develop during the second-to-third trimester. Histological studies demonstrated the growth of thalamocortical fibers (Kostovic and Jovanov-Milosevic, 2006), the ingrowth of cortical axons, and the emergence of corticocortical fibers with primary gyrification (Takahashi et al., 2014). These fiber pathways interconnect different cortical regions, forming complex brain networks (Sporns et al., 2005). Altered fetal brain structural connectivity are linked with neurodevelopmental abnormalities (Jakab et al., 2015), underscoring the importance of characterizing normative structural connectivity development.

Diffusion magnetic resonance imaging (dMRI) is a powerful neuroimaging tool for the noninvasive characterization of neuronal microstructure and brain injuries (Assaf et al., 2019). It also enables the reconstruction of white matter (WM) fiber pathways via tractography (Jeurissen et al., 2019). Several early studies have explored fetal brain WM development using ex vivo dMRI tractography of brain specimens (Huang et al., 2006, 2009; Takahashi et al., 2012). Recent advances in fast imaging and processing techniques have facilitated in utero studies of fetal brain WM fiber bundles (Hooker et al., 2020; Jaimes et al., 2020; Wilson et al., 2021) and spatiotemporal atlases depicting their development (Khan et al., 2019; Chen et al., 2022).

In addition, dMRI-based fiber tractography allows quantifications of whole-brain structural wiring patterns. By defining brain regions as nodes and traced streamlines as edges, a structural connectivity network (SCN) can be established. SCN topology can be quantitatively measured by graph theory metrics. Global efficiency (Eglob) quantifies the exchange of information across the entire network, reflecting network integration. Local efficiency (Eloc), clustering coefficient (CC), and modularity reflect network segregation. Eloc also quantifies a network's resistance to localized failures. Small-worldness (SW) characterizes a network with both high clustering and short paths, indicating balanced integration and segregation. Details on network measures are provided below, Network properties. Previous studies have revealed integration and segregation processes in prenatal brain SCN using preterm neonatal (Brown et al., 2014; Zhao et al., 2019a) and postmortem fetal brain data (Song et al., 2017), highlighting increases in Eglob, Eloc, and clustering coefficient (CC), and the emergence of small-world and rich-club organizations (Ball et al., 2014; Brown et al., 2014; Song et al., 2017; Zhao et al., 2019a). Fetal brain functional connectivity network (FCN) studies also indicated segregation patterns such as increasing CC (Cao et al., 2017), the existence of network communities (Turk et al., 2019), and integration patterns involving strengthened intermodule connections (Thomason et al., 2014). Complex functional topologies such as small-world, rich-club, and functional gradients were also observed (Fransson et al., 2011; Thomason et al., 2014; Turk et al., 2019; Moore et al., 2023).

While some studies used in utero fetal brain dMRI data to construct SCNs (Jakab et al., 2015; Hunt et al., 2018), they focused on specific conditions or comparative analyses, lacking descriptions of temporal changes or integration–segregation tendencies in the normal fetal brain SCN. In this work, we hypothesized the existence of a balanced integration and segregation process during the in utero development of fetal brain SCN and tested this hypothesis by measuring the developmental trajectories of network properties indicative of these tendencies.

Exploring the asymmetry of the developing brain network is another intriguing area. Brain structural asymmetry offers insights into hemisphere-specific functions (Toga and Thompson, 2003). Previous research suggested stronger left hemisphere connectivity in neonatal brains (Ratnarajah et al., 2013), especially relevant for language and motor skills. Resting-state functional MRI highlighted stronger connections in the left language and auditory networks in fetuses between 18 and 37 gestational weeks (W), while the visual and attention networks exhibited rightward lateralization (Kasprian et al., 2011). However, the lateralization patterns of dMRI-based fetal brain SCN were not investigated.

This study aims to establish SCN using in utero fetal brain dMRI data from 26W to 38W. We focused on intrahemispheric corticocortical SCN and seek to assess the coexistence of network integration and segregation, characterize regional variations, and delineate early lateralization patterns in the fetal brain SCN.

Materials and Methods

Experimental design

In this study, we constructed SCN of the fetal brain using in utero dMRI data acquired from 26W to 38W (see below, dMRI data processing and Building connectome). Graph theory analysis was used to evaluate the network properties of the fetal brain SCN, such as the nodal strength and betweenness, global and local efficiency, modularity, shortest path length, clustering coefficient, and small-worldness (see below, Network properties). We then performed correlation analyses between these measures and GA to study the developmental changes in fetal brain SCN. In addition, we investigated the asymmetry of SCN and its developmental pattern (see below, Statistical analysis).

Subjects and data acquisition

We enrolled 171 healthy pregnant women at Shandong Provincial Hospital after obtaining approval from the Institutional Review Board. The enrollment criteria included the following: (1) gestational age (GA) between 24 and 38 weeks; (2) no maternal comorbidities; and (3) fetuses with no clinical, ultrasound, or MRI evidence of brain abnormalities. Exclusion criteria were as follows: (1) pregnant women with gestational diabetes mellitus or hypertensive disorder complicating pregnancy, (2) multiple pregnancies, and (3) fetuses with chromosomal or genetic abnormalities diagnosed by amniocentesis. The scans were performed on a 3 T MR scanner (MAGNETOM Skyra, Siemens Healthineers) using an 18-channel body coil. The pregnant women were in the supine position when the fetuses were being scanned. We adopted a diffusion-weighted echoplanar imaging sequence with a b-value of 600 s/mm2, 30 gradient directions, eight nondiffusion-weighted images, 1.73 mm × 1.73 mm in-plane resolution in axial slices, 4 mm slice thickness, and two averages. Other acquisition parameters were as follows: field of view (FOV), 260 × 260 mm2; repetition time (TR), 3,900 ms; and echo time (TE), 80 ms. GeneRalized Auto-calibrating Partial Parallel Acquisition (GRAPPA) parallel imaging with an acceleration factor of 2 was adopted. Acquisition time for each subject was 3–4 min. Quality checks were performed before and after data processing by visual inspections to exclude images with poor data quality. This ensured that only high-quality images were used in the study. The data inclusion and exclusion flowchart was shown in Figure 1A.

dMRI data processing

The dMRI data processing was conducted using MRtrix3 and SVRTK software. The raw data underwent a preprocessing pipeline in MRtrix3 (https://www.mrtrix.org/), including denoising, volume-to-volume alignment, correction of eddy-current–induced distortions, and bias field correction (Tournier et al., 2019). The images were further processed using SVRTK (https://github.com/SVRTK/SVRTK; Deprez et al., 2020). The pipeline first performed iterative slice-to-volume registration (SVR) and superresolution reconstruction of all 2D image slices, regardless of gradient directions, to generate a motion-free mean diffusion-weighted (DW) 3D volume at 1.2 mm isotropic resolution (Kuklisova-Murgasova et al., 2012), which was used to initialize the spherical harmonics model. The original DW slices along with their slice-specific motion parameters resulting from SVR were then fitted to reconstruct all DW volumes with the corrected gradient tables (Deprez et al., 2020).

Then the FODs were calculated using constrained spherical deconvolution in MRtrix3 (Tournier et al., 2007). Probabilistic tractography was performed based on the FODs using a signal cutoff of 0.08 and a maximum angle threshold of 30° (Tournier et al., 2010; Chen et al., 2022). Given the considerable variability in fetal brain sizes during the studied period, we randomly selected 10 seeds per voxel to initiate tracking, which ensured a uniform seeding density across different brains. Subsequently, we applied SIFT2 to the resulting tractograms to match the distribution of streamlines to the underlying FOD signals and mitigate false positives (Smith et al., 2015).

Building connectome

Thirty-nine cortical regions in each hemisphere were defined based on the CRL fetal brain parcellations (Gholipour et al., 2014). First, the b0 image of each subject was nonlinearly registered to the CRL fetal brain T2W atlas at the corresponding GA using the SyN registration algorithm in Advanced Normalization Tools (https://github.com/ANTsX/ANTs). Next, the parcellation in the atlas space was back-transformed to the individual subject spaces using the inverse transformations from the previous step. The nodes were restricted to the cortical plate, by combining the CRL parcellation with manually delineated tissue segmentation (Xu et al., 2022).

The edge strength between every two nodes i and j was defined by the weighted sum of the streamlines. Specifically, the total number of streamlines between noes i and j was denoted as Nij and Wn was the weight of the nth streamline derived by SIFT2. Since larger ROIs inherently contain more streamlines than smaller ones, which could introduce a bias in favor of nodes with larger volumes exhibiting stronger connectivity, we normalized the number of streamlines by the sum of volumes of the two nodes, which is particularly important in characterizing the developing SCN with growing ROI volumes. Thus, the connectivity between node i and j was expressed as follows:Cij=∑n=1NijWnVi+Vj,(1) (1) where Vi and Vj are the numbers of voxels in the ROIs. Due to the challenges associated with fetal brain dMRI data acquisition, the noisy signals might cause spurious connections during fiber tracking. To minimize the impact of false-positive connections on the edge-based analysis, we set connections with <20 streamlines as zeros. It is important to note that our study focused solely on an intrahemispheric analysis excluding the interhemispheric streamlines, due to the inevitable false-positive interhemispheric connections at early GAs (see Discussion, Challenges in in utero fetal brain SCN analysis). The pipeline for building connectome was shown in Figure 2.

Network analysis

Network properties

We calculated the global properties of each intrahemispheric network. Global efficiency (Eglob), mean local efficiency (Eloc), shortest path length (Lp), nodal strength, and nodal betweenness were calculated in MATLAB using the GRETNA 2.0 toolbox (https://www.nitrc.org/projects/gretna/). Network efficiency assesses how effectively information flows in the network. Eglob and Eloc focus on the overall network and local groups, respectively. Lp measures the shortest route between nodes in a network. The network modularity evaluates how a network divides into distinct communities, which was calculated using the Louvain community detection algorithm with a resolution coefficient (gamma) of 2, evaluating how the networks divide into distinct communities. We adopted the algorithm proposed by Zhang and Horvath (2005) to calculate the clustering coefficient (CC) generalized for weighted networks, which was defined as follows:CCi=∑u,u≠i∑v,v≠i,v≠uaiuauvavi(∑u≠iaiu)2−∑u≠iaiu2,(2) where a is a soft adjacency matrix that ranks the nodes according to their connection strength and nodes u and v are the direct neighbors of node i (Zhang and Horvath, 2005). CC measures the extent to which nodes are clustered in a network. The small-worldness (SW) characterizes the networks that possess both high clustering like regular networks and short path lengths like random networks, which was calculated by the following:SW=CCnormLpnorm,(3) where CCnorm and Lpnorm are the network's CC and Lp normalized by the mean CC and Lp of 100 degree-matched random networks of the original network (Eq. 4).CCnorm=CCCCrandom,Lpnorm=LpLprandom.(4)

Statistical analysis

Statistical analyses were performed in MATLAB using the GRETNA 2.0 toolbox. To investigate the age dependence of fetal brain SCN, we used Pearson's correlation analysis to evaluate the change of network properties, nodal strength, nodal betweenness, and edge strength with GA. To study the asymmetry of fetal brain SCN, we used paired t tests to compare the network properties, nodal strength, nodal betweenness, and edge strength between the left and right hemispheres of each subject. Subject GA was defined as a covariate when making comparisons. To quantify the degree of laterality patterns, the laterality index (LI) of each measurement was calculated by the following:LI=100×R−LR+L,(5) where R and L are the corresponding network properties in the right and left brain hemispheres. LI > 0 represents a rightward lateralization and LI < 0 represents a leftward lateralization. The correlation between LI and GA was also calculated to evaluate the change of laterality over time. All statistical analyses with p values <0.05 after false discovery rate (FDR) correction were considered significant.

Results

Subjects

Out of 171 in utero dMRI scans acquired from 24W to 38W, 13 were excluded due to excessive signal dropouts from intraslice head motion. The remaining 158 scans underwent dMRI preprocessing and reconstruction, FOD estimation, and tractography. Twenty-six scans with signal bias that could not be corrected and 18 with poor fiber tracking results due to noisy reconstructed images were subsequently excluded. Ultimately, a total of 114 scans from 26 to 38 weeks were used for SCN analysis (Fig. 1). The sex of each fetus was unrevealed in compliance with the data privacy policy.⇓⇓

Figure 1.
Figure 1.

A, Flowchart of data inclusion and exclusion. B, The gestational age distribution of included fetal subjects.

Figure 2.
Figure 2.

The pipeline for building fetal brain structural connectivity network. Thirty-nine cortical regions in each hemisphere were defined based on the CRL atlas. Whole-brain probabilistic tractography was performed followed by SIFT2 streamline reweighting. The edge strength was calculated by the weighted sum of streamlines between every two nodes normalized by the number of voxels in nodal ROIs.

Figure 3.
Figure 3.

Change of network properties with age. Eglob, Eloc, and CC increased significantly and Lp decreased significantly over GA in both hemispheres. SW was found above 1 throughout the late-second-to-third gestational trimester and remained stable. The modularity had no significant correlation with GA. Eglob (t = 2.77; FDR-corrected p = 0.013), Eloc (t = 3.41; FDR-corrected p = 0.003), CC (t = 3.65; FDR-corrected p = 0.002), and SW (t = 2.60; FDR-corrected p = 0.016) were significantly lateralized toward the left hemisphere. Lp showed significant rightward lateralization (t = −2.37; FDR-corrected p = 0.023). The network modularity had no significant asymmetry (*p < 0.05, **p < 0.01, ***p < 0.001).

Network properties

Eglob, Eloc, and CC increased significantly while Lp decreased significantly over GA in both hemispheres. SW was found above 1 throughout the late-second-to-third gestational trimester and remained stable. The modularity had no significant correlation with GA (Fig. 3).

Regional connectivity patterns

Nodal strength exhibited widespread significant increase with GA in (FDR-corrected p < 0.05; Fig. 4A), with the correlation coefficient ranging from r = 0.24 (in the left Fusiform) to r = 0.91 (in the left medial orbital part of the superior frontal gyrus, ORBmed). We identified the nodes with r values exceeding the mean plus one standard deviation of all nodal r values as those with the most significant nodal strength increase. Notably, the bilateral ORBmed, anterior cingulate gyrus (ACG), and middle cingulate gyrus (MCG) and the right calcarine and MOG displayed the most significant growth in nodal strength (with r ranging from 0.77 to 0.91). Conversely, the degree of the right parahippocampal gyrus showed a negative correlation with GA (r = −0.31; FDR-corrected p = 0.001).

Figure 4.
Figure 4.

Regional development of fetal brain SCN. A, Cortical regions with a significant change in nodal strength with GA. B, Cortical regions with a significant change in nodal betweenness with GA. C, Edges with significant increase in strength over GA.

A significant increase in nodal betweenness was mainly observed in the parietal, middle temporal, and cingulate cortices (Fig. 4B), with the correlation coefficient ranging from r = 0.22 (in the right olfactory) to r = 0.47 (in the right anterior cingulum gyrus). The bilateral insula, cuneus, parahippocampal gyrus, and the right rolandic operculum (ROL) exhibited a significant decrease in nodal betweenness over time (with r ranging from −0.24 to −0.51).

Edges were observed to be significantly strengthened over GA, with the most significant increase occurring in the short-ranged connections in the inferior frontal cortex and around the central sulcus (Fig. 4C). A few long-ranged edges were found to be negatively correlated with GA (Fig. 6B), which however, may suffer from errors induced by the overestimation of WM connectivity in early GAs (discussed in Discussions, Challenges in in utero fetal brain SCN analysis).

Lateralization

In terms of global network metrics, Eglob (t = 2.77; FDR-corrected p = 0.013), Eloc (t = 3.41; FDR-corrected p = 0.003), CC, and SW (t = 2.60; FDR-corrected p = 0.016) were significantly lateralized toward the left hemisphere. Lp showed significant rightward lateralization (t = −2.37; FDR-corrected p = 0.023). The network modularity displayed no significant asymmetry. The LI of the global network properties did not exhibit any significant correlation with GA.

Regarding nodal betweenness, the orbital part of the inferior frontal gyrus (ORBinf), the olfactory, the angular gyrus, ROL, and cuneus exhibited significant lateralization to the left (Fig. 5A), with the ORBinf displaying increasing leftward lateralization and the olfactory showing decreasing leftward lateralization with GA (Fig. 5D). In contrast, the rectus, posterior cingulate, inferior and superior occipital, and middle and superior temporal gyri (STG), had significantly higher betweenness in the right hemisphere (Fig. 5A). The LI of the supramarginal gyrus increased significantly with GA (Fig. 5D) but this area had no significant lateralization in general.

Figure 5.
Figure 5.

Regional lateralization of fetal brain SCN. A–C, Lateralization of nodal betweenness, nodal strength, and edge strength. Inferior frontal and anterior-to-superior temporal nodes exhibited the most significant leftward lateralization while the posterior temporal and the occipital cortices had rightward lateralization. D–F, Correlation between LI of the regional properties (nodal betweenness, nodal strength, and edge strength) and GA. ANG, angular gyrus; CAL, calcarine; CUN, cuneus; HES, Heschl's gyrus; IFGoperc, opercular part of the inferior frontal gyrus; IFGtri, triangular part of the inferior frontal gyrus; IOG, inferior occipital gyrus; ITG, inferior temporal gyrus; MCG, middle cingulate gyrus; MFG, middle frontal gyrus; MOG, middle occipital gyrus; MTG, middle temporal gyrus; OLF, olfactory; ORBinf, orbital part of the inferior frontal gyrus; ORBmed, medial orbital part of the superior frontal gyrus; PCG, posterior cingulate gyrus; PCL, paracentral lobule; PCUN, precuneus; ROL, rolandic operculum; SFG, superior frontal gyrus; SFGmed, medial part of the superior frontal gyrus; SMA, supplementary motor area; SMG, supramarginal gyrus; SOG, superior occipital gyrus; STG, superior temporal gyrus; TPOsup, superior temporal pole.

The nodal strength, which reflects the strength of connectivity for each node in the network, demonstrated leftward lateralization in temporal regions including the superior temporal pole, ROL, inferior temporal, and Heschl's gyri. Frontal regions with leftward asymmetry involved the ORBinf, ORBmed, the triangular part of the inferior frontal gyrus (IFGtri), the middle frontal gyrus, and the supplementary motor area (SMA). The precuneus and cuneus also showed leftward asymmetry (Fig. 5B). The LI of nodal strength had a significant negative correlation with GA in ROL (Fig. 5E), which means it had increasing leftward lateralization. Nodes showing significant rightward lateralization primarily located in the occipital, parietal, and limbic cortices, including the calcarine, cingulate, supramarginal, inferior and superior occipital gyri, and paracentral lobule (Fig. 5B).

The distribution of lateralized network edges, which also rely on the strength of connectivity, showed similar spatial patterns as nodal strength. Edges with leftward asymmetry were observed in all brain lobes, while rightward-lateralized edges were mostly distributed in the posterior brain (Fig. 5C). LI of a few edges showed a significant correlation with GA. LI of the edges between ROL and the frontal cortex, between SMA and the paracentral lobule, as well as between the inferior temporal gyrus and middle occipital gyrus showed a negative correlation with GA. LI of the edge between the inferior occipital and middle temporal gyri, between the supramarginal gyrus and STG, and between the opercular part of the inferior frontal gyrus and the medial part of the superior frontal gyrus had a positive correlation with GA (Fig. 5F).

Discussion

In this study, we collected in utero dMRI to evaluate the normal developmental trajectory of the fetal brain SCN. The observed increases in Eglob, Eloc, and CC, decrease in Lp, and persistent SW from 26W to 38W reflected balanced network integration and segregation, supported by region-specific changes in nodal strength, betweenness, and edge strength. Eglob, Eloc, CC, and SW exhibited significant leftward lateralization, while Lp showed rightward asymmetry. The laterality of nodal strength, betweenness, and edge strength varied across brain regions. Our results provided valuable insights into fetal brain SCN development during the second-to-third trimester.

The normal development of fetal brain SCN

Development of network topology

The increasing Eglob and Eloc with GA indicated enhanced global and local capacities. The decreasing Lp suggested nodes becoming more closely connected. Increase in CC indicated enhancing local clustering. The small-worldness persisted throughout the studied period, reflecting both high clustering and short paths in the network, maintaining high capacity for both local and global information transfer, consistent with previous findings in fetuses (Marami et al., 2017; Song et al., 2017). Our observations validated the balance between network integration and segregation during in utero development (Zhao et al., 2019b), a distinction from postnatal network integration (Huang et al., 2015).

Regional reorganization

The regional analyses revealed general strengthened network connectivity. The growth rate of nodal strength, implied by their correlations with GA, exhibited posterior-to-anterior and medial-to-lateral gradients, which were the most significant in the bilateral ORBmed, ACG, and MCG, along with the right calcarine and SOG. The rapid growth of the calcarine and SOG supported prior findings that the primary visual network strengthened significantly in preterm infants (Cao et al., 2017). The ORBmed is a neonatal structural hub (van den Heuvel and Sporns, 2013) and becomes a functional hub in 2-year-old infants, indicating the emerging default mode network (Gao et al., 2011). The ACG and MCG were reported to be sensitive to infant crying (Mutschler et al., 2016), similar to their social functioning in adults (Fujiwara et al., 2007). ACG also becomes a structural hub in toddlerhood (Huang et al., 2015). These postnatal patterns might be supported by the rapid prenatal structural development we observed.

The strengthening of short-ranged edges coincided with the emergence of corticocortical u-fibers during the second-to-third trimester (Takahashi et al., 2012), possibly contributing to the segregation pattern revealed by increasing Eloc and CC. A few long-ranged edges, negatively correlated with GA, may stem from the overestimations of fiber connectivity in early GAs (Fig. 6B).

Figure 6.
Figure 6.

Unique challenges in fetal brain SCN analyses using in utero dMRI. A, Colored FA maps and whole-brain tractography in the 26W and 36W fetal brains. The bilateral frontal cortex “touched” with each other on the 26W colored FA and tractography due to insufficient resolution. B, Edges with significant negative correlation with GA. C, The cingulum bundle on the colored FA maps and in tractography. It was blurred with cortical fiber structures on the 26W fetal brain MRI (yellow arrow) and appeared thicker in tractography.

Nodal betweenness notably increased in bilateral frontal, temporal, and cingulate cortices, suggesting their increasingly central roles in the brain, influencing the network's shortest paths. These observations collectively implied intricate reorganization within the fetal brain SCN, characterized by strengthened local connections and a restructured centrality of nodes, maintaining an efficient small-world network organization.

Lateralization

Global lateralization patterns

The Eglob, Eloc, CC, and SW showed significant leftward lateralization and Lp was significantly lower in the left hemisphere, indicating that the left fetal brain hemisphere exhibits closer connections between nodes and within neighborhoods. These observations suggested higher information-transferring capacity in the left hemisphere of the fetal brain, similar to patterns in infants and right-handed adults (Ratnarajah et al., 2013; Caeyenberghs and Leemans, 2014). Our findings suggested that the global laterality patterns are innate and have emerged during in utero neurodevelopment, independent of postnatal experience.

Regional lateralization and possible functional explanations

Regional analysis revealed varied lateralization patternsin the fetal brain. As task-based experiments in utero are challenging, we lack specific knowledge about regional fetal brain functions. Nevertheless, we can still compare the observed structural asymmetries with known functional distributions in later life. Specifically, we compared the observed patterns in the language, motor, and visual-related brain regions with previous studies. Overall, our findings highlighted strong similarities between fetal brain structural asymmetry and the known lateralization patterns in both structural and functional aspects.

Language

Leftward asymmetries in nodal strength or betweenness were observed in language-related regions including the angular, Heschl's, inferior temporal gyri, the superior temporal pole, ROL, and IFGtri, which comprises Broca's area. The betweenness of ORBinf, the degree of ROL, and edges connected with ROL exhibited increasing leftward lateralization. These observations suggested leftward structural lateralization in language-related regions during fetal stage, strengthening with age, similar to findings in neonates (Ratnarajah et al., 2013) and adults (Caeyenberghs and Leemans, 2014). It also coincided with resting-state functional patterns that regions potentially develop into Broca's area became more left-lateralized in older fetuses (Thomason et al., 2014).

Contrarily, the STG, partially overlapping with Wernicke's speech area, displayed rightward-lateralized nodal betweenness. Its leftward functional lateralization was observed in neonates (Perani et al., 2011). However, studies highlighted that cortical folding, associated with local corticocortical fiber connections (Takahashi et al., 2012), progresses earlier and faster in the right STG compared with the left (Dubois et al., 2008; Glasel et al., 2011; Habas et al., 2012; Williams et al., 2021; Xu et al., 2022). This morphological disparity likely induces preferential connectivity underlying the right STG, leading to the observed rightward lateralization.

Motor

Nodal strength displayed leftward asymmetry in SMA, part of the motor cortex (Nachev et al., 2008). The edge between SMA and another motor-related region, the paracentral lobule, showed leftward change with GA (Fig. 4F), possibly leading to the leftward asymmetry of motor-related regions in later development (Ratnarajah et al., 2013; Caeyenberghs and Leemans, 2014).

Visual

Rightward lateralization was observed in primary visual cortices and visuospatial-related regions including the calcarine, cingulate, middle temporal, superior, and inferior occipital gyri. This agreed with the rightward functional asymmetry in the primary and association visual cortices reported in neonates (Williams et al., 2023). Among these regions, the nodal strength of the right calcarine and SOG had faster growth with GA compared with others, resembling the increasing nodal efficiency with age in visuospatial-related regions in the right hemisphere of adults (Caeyenberghs and Leemans, 2014).

To summarize, our observations provided insights into the structural lateralization of the fetal brain and their potential implications for functional laterality in language, motor, and visual-related regions. The specific relationship between structural and functional connectivity needs further investigation.

Challenges in in utero fetal brain SCN analysis

Our study provided preliminary insights into the developmental features of fetal brain SCN using in utero dMRI-based tractography. However, unique features of in utero fetal brain dMRI data present considerable challenges that warrant attention in future investigations. This section summarizes these challenges, particularly in relation to limited image quality and the rapid fetal brain development, and provides suggestions for future investigations.

Immaturity of fetal brain white matter

The fetal brain exhibits tissue heterogeneity and ongoing maturation processes, impacting dMRI properties and complicating tractography. In fetal brains younger than 30W, the highly anisotropic radial glial cells (RGCs) can be estimated by tractography but are not true WM fibers (Kostovic et al., 2002). The RGCs guide neuronal migration and gradually diminish as the process concludes (Sidman and Rakic, 1973; McKinstry et al., 2002). The weakened RGC-related connectivity with GA does not signify a decline in WM connectivity (Fig. 6B), warranting cautious interpretation of fetal brain connectivity from dMRI tractography.

Limited image quality

Acquiring high-quality in utero fetal brain dMRI data is hindered by the unpredictable fetal motion and insufficient resolution relative to the small brain sizes, introducing artifacts and partial volume effects. In our results, for example, the bilateral frontal cortex “touched” with each other (Fig. 6A), and the cingulum bundle merged with adjacent fibers (Fig. 6C). These effects led to overestimations of structural connectivity in early GAs, especially in the interhemispheric regions. To ensure reliability of our findings, we excluded interhemispheric data and performed separate calculations within each hemisphere. This approach allowed investigations on specific scientific questions such as lateralization but overlooked interhemispheric connections. Therefore, future studies aiming for comprehensive fetal brain SCN analyses with higher image quality should integrate interhemispheric connections. Additionally, including subcortical structures can enrich our understanding of the complete fetal brain connectome.

Moreover, employing low imaging resolutions might overlook short-ranged connections, potentially underestimating network measures such as local efficiency. Future studies need sufficient imaging resolution to address this bias.

Handling fetal brains of different GAs

Another inherent challenge in fetal brain imaging studies lies in the considerable difference in brain sizes across different GAs. Employing a uniform imaging resolution, while ensuring standardized comparisons, presents challenges in capturing detailed structures in smaller fetal brains. Using adaptive imaging resolutions for different brains could potentially establish a common “effective resolution,” yet it might yield disparate observations that require careful interpretation. Balancing between keeping methodological consistency and accommodating anatomical variation is crucial.

As for tractography, our study applied a common threshold for all subjects to eliminate erroneous connections, similar to Song et al. (2017). While prevalent studies adopted uniform tracking strategies for fetal brains of different GAs (Hooker et al., 2020; Machado-Rivas et al., 2021), different seeding densities and signal cutoffs across GAs can offer insights into development-related SCN changes. Moreover, the choices of these values lack gold standards and should be rigorously tested when dealing with different datasets.

Conclusion

In this study, we utilized in utero dMRI data collected from 26 to 38 weeks of gestation to investigate the developmental trajectories of the fetal brain structural connectivity network. Our analysis revealed significant increases in both global efficiency and mean local efficiency during the studied period. Small-worldness was found as early as 26 weeks and remained stable until birth. These developmental patterns reflected balanced network integration and segregation in the fetal brain. Short-ranged corticocortical connectivity strengthened significantly, and the development of nodal strength displayed a posterior-to-anterior and medial-to-lateral developmental sequence. The network efficiency and small-worldness showed significant lateralization toward the left hemisphere. The identified lateralization patterns in language, motor, and visual-related regions were consistent with known lateralization patterns in brain structure and function. An exception was observed in Wernicke's area, which displayed a contrasting rightward structural lateralization compared with its known leftward functional asymmetry. Overall, our findings provide novel insights into the development of the fetal brain network and its lateralization.

Footnotes

  • This work was supported by the Ministry of Science and Technology of the People’s Republic of China (2021ZD0200202), the National Natural Science Foundation of China (81971606, 82122032), the Science and Technology Department of Zhejiang Province (202006140, 2022C03057), and the Young Investigator Program of United Imaging (UIH-QNJJ-2021001).

  • The authors declare no competing financial interests.

  • Correspondence should be addressed to Dan Wu at danwu.bme{at}zju.edu.cn or Guangbin Wang at wgb7932596{at}hotmail.com.

References

  1. ↵
    1. Assaf Y,
    2. Johansen-Berg H,
    3. de Schotten MT
    (2019) The role of diffusion MRI in neuroscience. NMR Biomed 32:e3762. https://doi.org/10.1002/nbm.3762
    OpenUrl
  2. ↵
    1. Ball G, et al.
    (2014) Rich-club organization of the newborn human brain. Proc Natl Acad Sci U S A 111:7456–7461. https://doi.org/10.1073/pnas.1324118111 pmid:24799693
    OpenUrlAbstract/FREE Full Text
  3. ↵
    1. Brown CJ,
    2. Miller SP,
    3. Booth BG,
    4. Andrews S,
    5. Chau V,
    6. Poskitt KJ,
    7. Hamarneh G
    (2014) Structural network analysis of brain development in young preterm neonates. Neuroimage 101:667–680. https://doi.org/10.1016/j.neuroimage.2014.07.030
    OpenUrlCrossRefPubMed
  4. ↵
    1. Caeyenberghs K,
    2. Leemans A
    (2014) Hemispheric lateralization of topological organization in structural brain networks. Hum Brain Mapp 35:4944–4957. https://doi.org/10.1002/hbm.22524 pmid:24706582
    OpenUrlCrossRefPubMed
  5. ↵
    1. Cao M, et al.
    (2017) Early development of functional network segregation revealed by connectomic analysis of the preterm human brain. Cereb Cortex 27:1949–1963. https://doi.org/10.1093/cercor/bhw038 pmid:26941380
    OpenUrlCrossRefPubMed
  6. ↵
    1. Chen R,
    2. Sun C,
    3. Liu T,
    4. Liao Y,
    5. Wang J,
    6. Sun Y,
    7. Zhang Y,
    8. Wang G,
    9. Wu D
    (2022) Deciphering the developmental order and microstructural patterns of early white matter pathways in a diffusion MRI based fetal brain atlas. Neuroimage 264:119700. https://doi.org/10.1016/j.neuroimage.2022.119700
    OpenUrlCrossRef
  7. ↵
    1. Deprez M, et al.
    (2020) Higher order spherical harmonics reconstruction of fetal diffusion MRI with intensity correction. IEEE Trans Med Imaging 39:1104–1113. https://doi.org/10.1109/TMI.2019.2943565
    OpenUrlCrossRefPubMed
  8. ↵
    1. Dubois J,
    2. Benders M,
    3. Cachia A,
    4. Lazeyras F,
    5. Leuchter RH-V,
    6. Sizonenko SV,
    7. Borradori-Tolsa C,
    8. Mangin JF,
    9. Hueppi PS
    (2008) Mapping the early cortical folding process in the preterm newborn brain. Cereb Cortex 18:1444–1454. https://doi.org/10.1093/cercor/bhm180
    OpenUrlCrossRefPubMed
  9. ↵
    1. Fransson P,
    2. Aden U,
    3. Blennow M,
    4. Lagercrantz H
    (2011) The functional architecture of the infant brain as revealed by resting-state fMRI. Cereb Cortex 21:145–154. https://doi.org/10.1093/cercor/bhq071
    OpenUrlCrossRefPubMed
  10. ↵
    1. Fujiwara H,
    2. Hirao K,
    3. Namiki C,
    4. Yamada M,
    5. Shimizu M,
    6. Fukuyama H,
    7. Hayashi T,
    8. Murai T
    (2007) Anterior cingulate pathology and social cognition in schizophrenia: a study of gray matter, white matter and sulcal morphometry. Neuroimage 36:1236–1245. https://doi.org/10.1016/j.neuroimage.2007.03.068
    OpenUrlCrossRefPubMed
  11. ↵
    1. Gao W,
    2. Gilmore JH,
    3. Giovanello KS,
    4. Smith JK,
    5. Shen D,
    6. Zhu H,
    7. Lin W
    (2011) Temporal and spatial evolution of brain network topology during the first two years of life. PLoS One 6:e25278. https://doi.org/10.1371/journal.pone.0025278 pmid:21966479
    OpenUrlCrossRefPubMed
  12. ↵
    1. Gholipour A,
    2. Limperopoulos C,
    3. Clancy S,
    4. Clouchoux C,
    5. Akhondi-Asl A,
    6. Estroff JA,
    7. Warfield SK
    (2014) Construction of a deformable spatiotemporal MRI atlas of the fetal brain: evaluation of similarity metrics and deformation models. In: Medical image computing and computer-assisted intervention - MICCAI 2014 (Golland P, Hata N, Barillot C, Hornegger J, Howe R, eds) Vol. 8674, pp 292–299. Cham: Springer. https://doi.org/10.1007/978-3-319-10470-6_37 pmid:25485391
    OpenUrlPubMed
  13. ↵
    1. Glasel H,
    2. Leroy F,
    3. Dubois J,
    4. Hertz-Pannier L,
    5. Mangin JF,
    6. Dehaene-Lambertz G
    (2011) A robust cerebral asymmetry in the infant brain: the rightward superior temporal sulcus. Neuroimage 58:716–723. https://doi.org/10.1016/j.neuroimage.2011.06.016
    OpenUrlCrossRefPubMed
  14. ↵
    1. Habas PA,
    2. Scott JA,
    3. Roosta A,
    4. Rajagopalan V,
    5. Kim K,
    6. Rousseau F,
    7. Barkovich AJ,
    8. Glenn OA,
    9. Studholme C
    (2012) Early folding patterns and asymmetries of the normal human brain detected from in utero MRI. Cereb Cortex 22:13–25. https://doi.org/10.1093/cercor/bhr053 pmid:21571694
    OpenUrlCrossRefPubMed
  15. ↵
    1. Hooker JD, et al.
    (2020) Third-trimester in utero fetal brain diffusion tensor imaging fiber tractography: a prospective longitudinal characterization of normal white matter tract development. Pediatr Radiol 50:973–983. https://doi.org/10.1007/s00247-020-04639-8
    OpenUrlCrossRefPubMed
  16. ↵
    1. Huang H, et al.
    (2006) White and gray matter development in human fetal, newborn and pediatric brains. Neuroimage 33:27–38. https://doi.org/10.1016/j.neuroimage.2006.06.009
    OpenUrlCrossRefPubMed
  17. ↵
    1. Huang H, et al.
    (2015) Development of human brain structural networks through infancy and childhood. Cereb Cortex 25:1389–1404. https://doi.org/10.1093/cercor/bht335 pmid:24335033
    OpenUrlCrossRefPubMed
  18. ↵
    1. Huang H,
    2. Xue R,
    3. Zhang J,
    4. Ren T,
    5. Richards LJ,
    6. Yarowsky P,
    7. Miller MI,
    8. Mori S
    (2009) Anatomical characterization of human fetal brain development with diffusion tensor magnetic resonance imaging. J Neurosci 29:4263–4273. https://doi.org/10.1523/JNEUROSCI.2769-08.2009 pmid:19339620
    OpenUrlAbstract/FREE Full Text
  19. ↵
    1. Hunt D,
    2. Dighe M,
    3. Gatenby C,
    4. Studholme C
    (2018) Comparing diffusion tensor and spherical harmonic tractography for in utero studies of fetal brain connectivity. In: Medical imaging 2018: biomedical applications in molecular, structural, and functional imaging (Gimi B, Krol A, eds) Vol. 10578, pp 1057809. Houston, TX: SPIE. https://doi.org/10.1117/12.2294476
    OpenUrl
  20. ↵
    1. Jaimes C,
    2. Machado-Rivas F,
    3. Afacan O,
    4. Khan S,
    5. Marami B,
    6. Ortinau CM,
    7. Rollins CK,
    8. Velasco-Annis C,
    9. Warfield SK,
    10. Gholipour A
    (2020) In vivo characterization of emerging white matter microstructure in the fetal brain in the third trimester. Hum Brain Mapp 41:3177–3185. https://doi.org/10.1002/hbm.25006 pmid:32374063
    OpenUrlCrossRefPubMed
  21. ↵
    1. Jakab A,
    2. Kasprian G,
    3. Schwartz E,
    4. Gruber GM,
    5. Mitter C,
    6. Prayer D,
    7. Schoepf V,
    8. Langs G
    (2015) Disrupted developmental organization of the structural connectome in fetuses with corpus callosum agenesis. Neuroimage 111:277–288. https://doi.org/10.1016/j.neuroimage.2015.02.038
    OpenUrl
  22. ↵
    1. Jeurissen B,
    2. Descoteaux M,
    3. Mori S,
    4. Leemans A
    (2019) Diffusion MRI fiber tractography of the brain. NMR Biomed 32:e3785. https://doi.org/10.1002/nbm.3785
    OpenUrl
  23. ↵
    1. Kasprian G,
    2. Langs G,
    3. Brugger PC,
    4. Bittner M,
    5. Weber M,
    6. Arantes M,
    7. Prayer D
    (2011) The prenatal origin of hemispheric asymmetry: an in utero neuroimaging study. Cereb Cortex 21:1076–1083. https://doi.org/10.1093/cercor/bhq179
    OpenUrlCrossRefPubMed
  24. ↵
    1. Khan S,
    2. Vasung L,
    3. Marami B,
    4. Rollins CK,
    5. Afacan O,
    6. Ortinau CM,
    7. Yang E,
    8. Warfield SK,
    9. Gholipour A
    (2019) Fetal brain growth portrayed by a spatiotemporal diffusion tensor MRI atlas computed from in utero images. Neuroimage 185:593–608. https://doi.org/10.1016/j.neuroimage.2018.08.030 pmid:30172006
    OpenUrlCrossRefPubMed
  25. ↵
    1. Kostovic I,
    2. Jovanov-Milosevic N
    (2006) The development of cerebral connections during the first 20-45 weeks’ gestation. Semin Fetal Neonatal Med 11:415–422. https://doi.org/10.1016/j.siny.2006.07.001
    OpenUrlCrossRefPubMed
  26. ↵
    1. Kostovic I,
    2. Judas M,
    3. Rados M,
    4. Hrabac P
    (2002) Laminar organization of the human fetal cerebrum revealed by histochemical markers and magnetic resonance imaging. Cereb Cortex 12:536–544. https://doi.org/10.1093/cercor/12.5.536
    OpenUrlCrossRefPubMed
  27. ↵
    1. Kuklisova-Murgasova M,
    2. Quaghebeur G,
    3. Rutherford MA,
    4. Hajnal JV,
    5. Schnabel JA
    (2012) Reconstruction of fetal brain MRI with intensity matching and complete outlier removal. Med Image Anal 16:1550–1564. https://doi.org/10.1016/j.media.2012.07.004 pmid:22939612
    OpenUrlCrossRefPubMed
  28. ↵
    1. Machado-Rivas F,
    2. Afacan O,
    3. Khan S,
    4. Marami B,
    5. Velasco-Annis C,
    6. Lidov H,
    7. Warfield SK,
    8. Gholipour A,
    9. Jaimes C
    (2021) Spatiotemporal changes in diffusivity and anisotropy in fetal brain tractography. Hum Brain Mapp 42:5771–5784. https://doi.org/10.1002/hbm.25653 pmid:34487404
    OpenUrlCrossRefPubMed
  29. ↵
    1. Marami B,
    2. Salehi SSM,
    3. Afacan O,
    4. Scherrer B,
    5. Rollins CK,
    6. Yang E,
    7. Estroff JA,
    8. Warfield SK,
    9. Gholipour A
    (2017) Temporal slice registration and robust diffusion-tensor reconstruction for improved fetal brain structural connectivity analysis. Neuroimage 156:475–488. https://doi.org/10.1016/j.neuroimage.2017.04.033 pmid:28433624
    OpenUrlCrossRefPubMed
  30. ↵
    1. McKinstry RC,
    2. Mathur A,
    3. Miller JH,
    4. Ozcan A,
    5. Snyder AZ,
    6. Schefft GL,
    7. Almli CR,
    8. Shiran SI,
    9. Conturo TE,
    10. Neil JJ
    (2002) Radial organization of developing preterm human cerebral cortex revealed by non-invasive water diffusion anisotropy MRI. Cereb Cortex 12:1237–1243. https://doi.org/10.1093/cercor/12.12.1237
    OpenUrlCrossRefPubMed
  31. ↵
    1. Moore JW, et al.
    (2023) Gradient organisation of functional connectivity within resting state networks is present from 25 weeks gestation in the human fetal brain. bioRxiv.
  32. ↵
    1. Mutschler I,
    2. Ball T,
    3. Kirmse U,
    4. Wieckhorst B,
    5. Pluess M,
    6. Klarhofer M,
    7. Meyer AH,
    8. Wilhelm FH,
    9. Seifritz E
    (2016) The role of the subgenual anterior cingulate cortex and amygdala in environmental sensitivity to infant crying. PLoS One 11:e0161181. https://doi.org/10.1371/journal.pone.0161181 pmid:27560361
    OpenUrlPubMed
  33. ↵
    1. Nachev P,
    2. Kennard C,
    3. Husain M
    (2008) Functional role of the supplementary and pre-supplementary motor areas. Nat Rev Neurosci 9:856–869. https://doi.org/10.1038/nrn2478
    OpenUrlCrossRefPubMed
  34. ↵
    1. Perani D,
    2. Saccuman MC,
    3. Scifo P,
    4. Awander A,
    5. Spada D,
    6. Baldoli C,
    7. Poloniato A,
    8. Lohmann G,
    9. Friederici AD
    (2011) Neural language networks at birth. Proc Natl Acad Sci U S A 108:16056–16061. https://doi.org/10.1073/pnas.1102991108 pmid:21896765
    OpenUrlAbstract/FREE Full Text
  35. ↵
    1. Ratnarajah N,
    2. Rifkin-Graboi A,
    3. Fortier MV,
    4. Chong YS,
    5. Kwek K,
    6. Saw S-M,
    7. Godfrey KM,
    8. Gluckman PD,
    9. Meaney MJ,
    10. Qiu A
    (2013) Structural connectivity asymmetry in the neonatal brain. Neuroimage 75:187–194. https://doi.org/10.1016/j.neuroimage.2013.02.052 pmid:23501049
    OpenUrlCrossRefPubMed
  36. ↵
    1. Sidman RL,
    2. Rakic P
    (1973) Neuronal migration, with special reference to developing human brain - review. Brain Res 62:1–35. https://doi.org/10.1016/0006-8993(73)90617-3
    OpenUrlCrossRefPubMed
  37. ↵
    1. Smith RE,
    2. Tournier J-D,
    3. Calamante F,
    4. Connelly A
    (2015) SIFT2: enabling dense quantitative assessment of brain white matter connectivity using streamlines tractography. Neuroimage 119:338–351. https://doi.org/10.1016/j.neuroimage.2015.06.092
    OpenUrlCrossRefPubMed
  38. ↵
    1. Song L,
    2. Mishra V,
    3. Ouyang M,
    4. Peng Q,
    5. Slinger M,
    6. Liu S,
    7. Huang H
    (2017) Human fetal brain connectome: structural network development from middle fetal stage to birth. Front Neurosci 11:561. https://doi.org/10.3389/fnins.2017.00561 pmid:29081731
    OpenUrlCrossRefPubMed
  39. ↵
    1. Sporns O,
    2. Tononi G,
    3. Kötter R
    (2005) The human connectome: a structural description of the human brain. PLoS Comput Biol 1:e42. https://doi.org/10.1371/journal.pcbi.0010042 pmid:16201007
    OpenUrlCrossRefPubMed
  40. ↵
    1. Takahashi E,
    2. Folkerth RD,
    3. Galaburda AM,
    4. Grant PE
    (2012) Emerging cerebral connectivity in the human fetal brain: an MR tractography study. Cereb Cortex 22:455–464. https://doi.org/10.1093/cercor/bhr126 pmid:21670100
    OpenUrlCrossRefPubMed
  41. ↵
    1. Takahashi E,
    2. Hayashi E,
    3. Schmahmann JD,
    4. Grant PE
    (2014) Development of cerebellar connectivity in human fetal brains revealed by high angular resolution diffusion tractography. Neuroimage 96:326–333. https://doi.org/10.1016/j.neuroimage.2014.03.022 pmid:24650603
    OpenUrlCrossRefPubMed
  42. ↵
    1. Thomason ME, et al.
    (2014) Intrinsic functional brain architecture derived from graph theoretical analysis in the human fetus. PLoS One 9:e94423. https://doi.org/10.1371/journal.pone.0094423 pmid:24788455
    OpenUrlCrossRefPubMed
  43. ↵
    1. Toga AW,
    2. Thompson PM
    (2003) Mapping brain asymmetry. Nat Rev Neurosci 4:37–48. https://doi.org/10.1038/nrn1009
    OpenUrlCrossRefPubMed
  44. ↵
    1. Tournier JD,
    2. Calamante F,
    3. Connelly A
    (2007) Robust determination of the fibre orientation distribution in diffusion MRI: non-negativity constrained super-resolved spherical deconvolution. Neuroimage 35:1459–1472. https://doi.org/10.1016/j.neuroimage.2007.02.016
    OpenUrlCrossRefPubMed
  45. ↵
    1. Tournier JD,
    2. Calamante F,
    3. Connelly A
    (2010) Improved probabilistic streamlines tractography by 2nd order integration over fibre orientation distributions. Proceedings of the International Society for Magnetic Resonance in Medicine.
  46. ↵
    1. Tournier JD,
    2. Smith R,
    3. Raffelt D,
    4. Tabbara R,
    5. Dhollander T,
    6. Pietsch M,
    7. Christiaens D,
    8. Jeurissen B,
    9. Yeh C-H,
    10. Connelly A
    (2019) MRtrix3: a fast, flexible and open software framework for medical image processing and visualisation. Neuroimage 202:116137. https://doi.org/10.1016/j.neuroimage.2019.116137
    OpenUrl
  47. ↵
    1. Turk E, et al.
    (2019) Functional connectome of the fetal brain. J Neurosci 39:9716–9724. https://doi.org/10.1523/JNEUROSCI.2891-18.2019 pmid:31685648
    OpenUrlAbstract/FREE Full Text
  48. ↵
    1. van den Heuvel MP,
    2. Sporns O
    (2013) Network hubs in the human brain. Trends Cogn Sci 17:683–696. https://doi.org/10.1016/j.tics.2013.09.012
    OpenUrlCrossRefPubMed
  49. ↵
    1. Williams LZJ, et al.
    (2023) Structural and functional asymmetry of the neonatal cerebral cortex. Nat Hum Behav 7:942–955. https://doi.org/10.1038/s41562-023-01542-8
    OpenUrl
  50. ↵
    1. Williams LZJ,
    2. Fitzgibbon SP,
    3. Bozek J,
    4. Winkler AM,
    5. Dimitrova R,
    6. Poppe T,
    7. Schuh A,
    8. Makropoulos A,
    9. Cupitt J,
    10. O’Muircheartaigh J
    (2021) Structural and functional asymmetry of the neonatal cerebral cortex. bioRxiv, 2021.2010. 2013.464206.
  51. ↵
    1. Wilson S, et al.
    (2021) Development of human white matter pathways in utero over the second and third trimester. Proc Natl Acad Sci U S A 118:e2023598118. https://doi.org/10.1073/pnas.2023598118 pmid:33972435
    OpenUrlAbstract/FREE Full Text
  52. ↵
    1. Xu X,
    2. Sun C,
    3. Sun J,
    4. Shi W,
    5. Shen Y,
    6. Zhao R,
    7. Luo W,
    8. Li M,
    9. Wang G,
    10. Wu D
    (2022) Spatiotemporal atlas of the fetal brain depicts cortical developmental gradient in Chinese population. bioRxiv.
  53. ↵
    1. Zhang B,
    2. Horvath S
    (2005) A general framework for weighted gene co-expression network analysis. Stat Appl Genet Mol Biol 4:1–45. https://doi.org/10.2202/1544-6115.1128
    OpenUrlCrossRef
  54. ↵
    1. Zhao T, et al.
    (2019a) Structural network maturation of the preterm human brain. Neuroimage 185:699–710. https://doi.org/10.1016/j.neuroimage.2018.06.047 pmid:29913282
    OpenUrlCrossRefPubMed
  55. ↵
    1. Zhao T,
    2. Xu Y,
    3. He Y
    (2019b) Graph theoretical modeling of baby brain networks. Neuroimage 185:711–727. https://doi.org/10.1016/j.neuroimage.2018.06.038
    OpenUrlCrossRef
Back to top

In this issue

Email
Print
View Full Page PDF
Citation Tools
Respond to this article
Development of the Fetal Brain Corticocortical Structural Network during the Second-to-Third Trimester Based on Diffusion MRI
Ruike Chen, Ruoke Zhao, Haotian Li, Xinyi Xu, Mingyang Li, Zhiyong Zhao, Cong Sun, Guangbin Wang, Dan Wu
Journal of Neuroscience 17 July 2024, 44 (29) e1567232024; DOI: 10.1523/JNEUROSCI.1567-23.2024
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords

Responses to this article

Respond to this article

Jump to comment:

No eLetters have been published for this article.