ArticlePDF AvailableLiterature Review

The functional brain connectome of the child and autism spectrum disorders


Abstract and Figures

Brain connectomics is a relatively new field of research that maps the brain's large-scale structural and functional networks at rest. The connectome of the human brain develops progressively from early infancy to late adolescence, and this review describes the theory behind the concept and its applicability to studying the development and dynamics of brain networks through graph theoretical metrics. We also describe how the brain connectome concept could further our understanding of autism spectrum disorders (ASD) Conclusion: Further research into the functional child brain connectome concept could enhance our understanding of atypical brain connectivity patterns presumed to be linked to ASD.
dentification of cortical hubs and their corresponding functional networks in the adult and infant brain, as revealed by resting-state functional magnetic resonance imaging. The spatial distribution of the magnitude of the degree centrality measure – the number of edges connected to each node, defined here as 59 5 9 5 mm 3 voxels – is shown in panels A (adults) and B (infants). In adults, a high degree of centrality can be observed to reside foremost in the default mode network regions, including the precuneus/posterior cingulate cortex and the medial prefrontal cortex. In contrast, infants showed a high degree of centrality, which was most prominent in the primary sensory cortices. The degree of betweenness centrality, which is commonly used as an estimate for putative candidates of cortical hubs, is shown in panels C (adults) and D (infants). Similar to the spatial distribution of degree centrality, candidate hub regions showed a preference for the default network regions in the adult brain, whereas primary sensory regions dominated the spatial profile of degree betweenness in infants. Panel E (adults) and F (infants) show typical examples of the spatial pattern networks associated with candidates for cortical hubs. A seed region (green sphere in the figure) positioned in the precuneus/posterior cingulate cortex, and a corresponding seed-based correlation analysis, yielded a connectivity pattern that stretched between the brain regions within the default network regions in adults. Analogously, a seed positioned in the left primary sensorimotor area, a likely candidate for a hub region in infants (panels B and D), produced a robust connectivity with the bilateral primary sensory network. Figure adapted from Fransson et al. (28).
Content may be subject to copyright.
The functional brain connectome of the child and autism spectrum disorders
Katell Mevel
, Peter Fransson (
1.Laboratory for the Psychology of Child Development and Education (LaPsyD
E), CNRS UMR 8240, Sorbonne Paris Cit
e, GIP Cyceron, Universit
e de Caen Normandie,
e Paris Descartes, Paris, France
2.Department of Clinical Neuroscience, Karolinska Institutet, Stockholm, Sweden
Autism spectrum disorders, Child brain connectome,
Connectivity, Functional magnetic resonance
imaging, Resting state
Peter Fransson, Nobels v
ag 9, Department of Clinical
Neuroscience, Karolinska Institutet, SE-171 77,
Stockholm, Sweden.
Tel: +46 (0)8 524 800 00 |
4 February 2016; revised 5 April 2016;
accepted 24 May 2016.
Brain connectomics is a relatively new field of research that maps the brain’s large-scale
structural and functional networks at rest. The connectome of the human brain develops
progressively from early infancy to late adolescence, and this review describes the theory
behind the concept and its applicability to studying the development and dynamics of brain
networks through graph theoretical metrics. We also describe how the brain connectome
concept could further our understanding of autism spectrum disorders (ASD)
Conclusion: Further research into the functional child brain connectome concept could
enhance our understanding of atypical brain connectivity patterns presumed to be linked
to ASD.
Research that aims to map the human brain connectome
has gained prominence in the neuroscience literature. The
term brain connectome has commonly been coined as a
collective term to describe using neuroimaging tools to
delineate the structural and functional large-scale networks
in the human brain (1). In contrast to studies that use task-
related functional neuroimaging to examine the relation-
ship between functional brain organisation and cognitive
skills, brain connectomics aims to build comprehensive
maps of all functional and structural connections in the
brain. It does this within the spatial and temporal
limitations inherent for the imaging modalities that are
most commonly used, namely structural and functional
magnetic resonance imaging, high-resolution electro-
encephalograms and magnetoencephalograms, to build
comprehensive maps of all structural and functional
connections in the brain.
ADHD, Attention-deficit hyperactivity disorder; ASD, Autism
spectrum disorders; Au, Auditory; BOLD, Blood oxygenation
level-dependent; CO, Cingulo-opercular; DA, Dorsal attention;
DM, Default mode; fMRI, Functional magnetic resonance imag-
ing; FP, Fronto-parietal attention; ROI, Region of interest; Sa,
Saliency; SM, Sensorimotor; Sub, Subcortical; VA, Ventral
attention; Vis, Visual.
Key notes
Brain connectomics maps the brain’s large-scale struc-
tural and functional networks at rest.
This study describes the theory behind the concept, the
development and dynamics of brain networks using
graph theoretical metrics and how the concept could
further our understanding of autism spectrum disorders
We conclude that further research into the functional
child brain connectome concept could enhance our
understanding of atypical brain connectivity patterns
presumed to be linked to ASD.
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 1
Acta Pædiatrica ISSN 0803-5253
The endeavour to parcellate the human brain into its
constituents in the form of large-scale functional networks
has largely been driven by advances in acquiring and
processing resting-state fMRI data. Resting-state fMRI is
based on the observation that the fMRI signal during rest is
synchronised across anatomically widespread brain areas.
This implies that anatomically separated brain regions are
functionally connected with each other, a property that can
be examined by computing the strength of correlation
between the resting-state blood oxygenation level-depen-
dent (BOLD) fMRI signal intensity time courses extracted
between different brain regions (2). Initial reports con-
firmed the presence of resting-state networks in primary
sensorimotor areas, but subsequent studies such as Damoi-
seaux et al. (3) unequivocally showed that virtually all areas
of the brain, including higher associative cortices, are
included in large-scale resting-state networks. It is impor-
tant to note that the term resting state is somewhat
misleading, as resting-state networks are also present during
task conditions (4,5). Hence, the term resting-state net-
works is often interchangeably used with the term intrinsic
networks, which emphasises the fact that fMRI network
connectivity is driven by endogenous BOLD signal fluctu-
ations. In contrast to task-evoked fMRI, resting-state fMRI
has the additional advantage that it can be carried out
without the subjects following instructions. This has turned
it into an attractive tool for investigating functional brain
connectivity in populations that cannot normally be
scanned in an functional neuroimaging environment. In
particular, the use of resting-state fMRI to investigate brain
connectivity in paediatric populations has been of signifi-
cant interest in recent years.
This review of the connectomics of the child brain starts
with a brief background and description of the key aspects
of intrinsic, resting-state fMRI brain connectivity, followed
by a short introduction on how graph theory can be applied
to functional neuroimaging data. We describe the key graph
theoretical concepts that are commonly used to charac-
terise the large-scale human brain connectome and how
resting-state fMRI can be combined with graph theoretical
methods to map the functional connectome of the infant
brain. We then review attempts to map developmental
changes in the functional brain connectome during early
childhood and adolescence. This is followed by a discussion
on some of the key findings in terms of delineating atypical
developmental patterns in brain connectivity in patient
populations with neurodevelopmental disorders, with an
emphasis on autism spectrum disorder (ASD). We conclude
by providing a brief summary and a future outlook on
research into the functional brain connectome in the child.
Resting-state fMRI
Biswal et al. (6) were the first researchers to discover that
BOLD fMRI signals acquired during rest, without any overt
behaviour from the subject, contained signal changes that
were of neuronal origin and were not related to peripheral
physiology. This seminal paper showed that during rest, the
BOLD fMRI signal intensity time course in the left primary
sensorimotor area was correlated to the BOLD signal
intensity time course in the right primary sensorimotor
cortex across time. The initial discovery of resting-state
functional connectivity in the sensorimotor system was
followed by studies that showed similar resting-state syn-
chronisation of the BOLD fMRI signal across the hemi-
spheres for the auditory and visual primary cortices. Later
studies, including Damoiseaux et al. (3), confirmed these
initial reports and extended the finding of resting-state fMRI
networks in the brain to include associative cortices. Over the
last decade, experimental evidence has accumulated that
consistently points to the conclusion that the human brain is
parcellated at a large spatial scale into a functionally
consistent network that encompasses the cerebrum and the
cerebellum. Importantly, the resting-state BOLD signal
fluctuations that give rise to fMRI resting-state brain con-
nectivity slowly evolve over time, typically residing in the
0.010.1 Hz frequency range. Although empirical research
has unequivocally shown that the low-frequency signal
changes that constitute the fMRI resting-state signal are
related to changes in neuronal activity, their relation to the
underlying neurophysiological events is not fully understood
at the moment. The paper by Leopold et al. (7) is relevant
here. The relationship between the resting-state BOLD fMRI
signal and electrophysiological signals in different frequency
bands, for example theta, alpha, beta and gamma frequency
band brain activity, is an active area of research and outside
the scope of the present review. Literature that provides an
in-depth narrative on this important topic includes Shmuel
et al. (8), Sch
olvinck et al. (9) and Chang et al. (10).
A model of the large-scale functional brain connectome
With the aid of resting-state fMRI and other noninvasive
functional neuroimaging techniques, the overarching aim of
brain connectomics is to provide researchers and clinicians
with an exhaustive map of all the functional connections in
the human brain. A further research goal is to develop a
description of the functional topology of the brain’s
networks that is sufficiently accurate and detailed to
contribute novel insights into the complex patterns of the
functional disabilities and impairments that often arise as a
consequence of neurological, neurodevelopmental and
psychiatric disorders (11,12). It has also become evident
that graph theory, a branch of mathematics, can be used to
great advantage to describe and examine key properties of
the brain connectome, such as integration and segregation
of brain function in the context of interacting networks.
There are a number of key concepts within graph theory
that are of central importance to the study of functional
brain connectivity based on resting-state fMRI data.
At its most basic level, a network consists of two nodes
region of interests (ROI) that represent two different
anatomical regions in the brain and an edge that represents
the presence of connectivity between any pair of nodes
(Fig. 1). Although brain networks can be studied in both the
structural and functional domains, we will mostly concern
ourselves with functional brain networks. Typically, an edge
between two nodes is said to exist if the corresponding
2©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
pair-wise correlation coefficient exceeds a given threshold
value. Naturally, depending on the spatial limitations of the
neuroimaging method used, the network model of the brain
may be augmented with an almost arbitrary number of
nodes and their corresponding edges, as shown in
Figure 1B. Once we have established which nodes and
brain regions we are interested in and formed a network
model of brain connectivity, we can start to formulate
questions about patterns of brain connectivity that can be
framed in a graph theoretical context. For example, we may
ask whether the topology of the functional brain connec-
tivity conforms to the small-world architecture shown on
the bottom left of Figure 1C or if the configuration of brain
connectivity is influenced by a given task, as shown on the
bottom right.
So how can a basic graph theoretical model be applied in
the case of functional brain connectivity measured by fMRI
during rest, that is resting-state fMRI brain connectivity?
The major steps involved in this process are outlined in
Figure 2. For the example shown in this figure, the brain is
parcellated into 264 nodes that are distributed across the
cortex, subcortical regions and the cerebellum as suggested
by Power et al. (13) (Fig. 2A). Each node is defined as a
spherical ROI with a 5-mm radius and the colour coding of
the nodes shown in Figure 2A indicates the membership of
previously reported resting-state subnetworks in the human
brain (13) and further details of the subnetworks are
provided in the legend. Once the anatomical localisation
of the nodes of interest has been determined, the mean
BOLD fMRI image intensity time series is extracted from all
image voxels belonging to the ROI (Fig. 2B). Because most
of the signal variance in the resting-state fMRI signal that is
of a neuronal origin resides in the low-frequency range, the
extracted ROI signal intensity time courses are typically
band-pass filtered. Hence, non-neuronal signal contribu-
tions are regressed out from the signal intensity time courses
before further analysis. These could come from, white
matter, cerebrospinal fluid and residual variance related to
subject micro-head movement and physiological noise
cardiac and respiratory pulsations, for example. Next, the
correlation coefficient ris computed between all pair-wise
combinations of ROIs and subsequently inserted into a
correlation matrix as shown in Figure 2C. By definition, this
correlation matrix is symmetrical, because the correlation
coefficients do not take directionality into account. In our
case, with a 264-node network parcellation of the brain
(Fig. 2A), the resulting correlation matrix will contain 264
rows and 264 columns. Typically, the correlation matrix is
Figure 1 Schematic figure that illustrates the concept of defining brain networks based on neuroimaging data. The underlying logic for the definition of networks in the
brain can be formulated in both the structural and functional domains. A structural connection between two arbitrary brain regions or equivalent nodes, A and B, is said
to be present if an anatomical connection exists between them (e.g. a white matter fibre tract detected with diffusion-weighted MRI) (A), whereas a functional connection
may exist if, for example, the corresponding resting-state blood oxygenation level-dependent functional magnetic resonance imaging signal intensities are correlated over
time (right). Note that the strength of connectivity, that is the amplitude of the correlation coefficient, is highlighted by the width of the edge that connects nodes A and B.
At the next level (B), we can define an ensemble of nodes that together constitute a network in which the strength of connectivity between individual nodes may vary
depending on the degree of structural (left) or functional connectivity (right). Once we have defined a network, we may, for example, investigate whether the distribution
of the structural connectivity edges conforms to a small-world architecture (C, left) or whether the strength of connectivity within the network is altered during different
tasks (C, right). Figure courtesy of William Hedley Thompson.
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 3
Mevel and Fransson The functional brain connectome of the child
then thresholded at some cut-off value, because it is not
biologically plausible that all nodes are connected to all
other nodes. By thresholding the correlation matrix, we are
left with a binary, unweighted connectivity matrix where
the occurrence of a value of one in the connectivity matrix
is interpreted as the presence of connectivity, namely an
edge, between two nodes (Fig. 2D). It should be noted that
it is also possible to assign each edge to have a value of
between zero and one, which would then constitute a
weighted connectivity matrix.
When a decision has been made about the granularity
and coverage of the brain network, that is the positioning
and number of the network nodes (Fig. 2A), and the nodes
that are considered to be correlated with each other (edges)
have been determined (Fig. 2C,D), the network model can
be analysed using graph theoretical methods to estimate
network properties that are biologically relevant. In this
review, we only discuss a few central network properties
that serve to illustrate the most important and basic
concepts. An extensive primer on graph theoretical mea-
sures and how they can be applied to network models of
large-scale brain connectivity can be found in the paper by
Rubinov and Sporns (14).
Functional integration and segregation in the brain from a
connectome perspective
Given the outline of functional brain connectivity described
in the previous section, we are now equipped with the
necessary background to describe the network properties of
the connectome that are biologically meaningful. In the
past, functional neuroimaging was mainly used to investi-
gate the segregation of cognitive function in terms of single
function structure relationships. However, a network-based
model of brain connectivity provides further possibilities
to assess the functional role of integration of neuronal
processes in the brain, as discussed in Fox and Friston (15).
An example of graph theoretical properties that are related
to these central concepts of brain function is given in
Figure 3A. In this simple network example with ten nodes,
two different properties are highlighted: node degree and
betweenness centrality. Node degree is defined as the
number of edges connected to each node, that is the
number of other brain regions that are significantly con-
nected with the node of interest. Typically, node degree is
associated with network centrality. Thus, nodes that rank
high on the centrality and degree scale, such as orange and
purple nodes, can be viewed as nodes that are interacting
with a higher number of nodes in the network than nodes
with a low centrality and degree, such as green nodes.
Nodes that have a high degree betweenness centrality are
often referred to as hubs, meaning that they might play a
pivotal role in the transfer of neuronal information within
the network. The number and spatial distribution of
functional hub nodes in the brain are thus essential for
the brain’s ability to be resilient towards injury. Network
resilience is often expressed as the vulnerability of the brain
Figure 2 Outline of the steps typically taken to compute measures of brain connectivity from resting-state functional magnetic resonance imaging (fMRI) data. We start
by defining a network of nodes (regions of interest) that covers the brain cortex as well as subcortical structures (A). In this example, the parcellation defined by Power
et al. (13) is used, which samples the cerebrum and cerebellum at 264 nodes that are divided into 10 separate resting-state subnetworks. These are colour-coded in the
figure according to default mode (DM) (brown), sensorimotor (SM) (orange), visual (Vis) (red), fronto-parietal attention (FP) (light blue), saliency (Sa) (light green),
cingulo-opercular (CO) (yellow), auditory (Au) (purple), subcortical network (sub) (magenta), dorsal attention (DA) (dark green) and ventral attention network (VA)
(cyan). (B) The blood oxygenation level-dependent fMRI resting-state signals (X
.. X
) from each node (spherical regions of interest, 5 mm radius) are extracted from
the functional fMRI images, and then, the pair-wise correlation coefficient (r) is computed. The pair-wise correlation coefficients for the signal intensity time courses are
used as measures of the strength of connectivity between any two nodes in the network. Subsequently, correlation coefficients are used to form the brain connectivity
matrix shown in panel C, where each element in the matrix signifies the strength of connectivity between any pair of nodes in the brain network. The presence of
connectivity between individual nodes in the networks can be visualised as edges between them as shown in panel D.
4©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
network to insults such as strokes and traumatic brain
injuries (11).
Brain integration refers to the brain’s capacity to organise
and extract information from brain areas that are dispersed
throughout the brain. To measure the brain’s ability to
integrate information from different brain regions and
subnetworks, we would typically invoke the concept of
paths in the network. A path can be described as a sequence
of nodes and edges in which neuronal information is routed
and flowed through the network. Thus, if we want to
estimate the degree of integration in the connectome, it
would be interesting to find the shortest paths between any
two nodes, as these imply stronger integration than pairs
that are separated by longer paths. A commonly used
measure to estimate integration in connectome models is
the betweenness centrality measure, which is defined as the
fraction of all the shortest paths in the connectome that
passes through a given node. Thus, nodes that connect to
many different subnetworks often score highly on the
betweenness centrality measure, as shown in purple in
Figure 3A. This particular node is connected to a relatively
large number of nodes, and it is also the only node that is
connected to the cluster nodes in the upper part of the
network (coloured in blue and green) as well as to the
cluster nodes in the lower part (coloured in orange and
brown). This means that any information that needs to be
routed from any of the blue or green nodes to any of the
orange or brown nodes must be routed through the purple-
coloured node. Hence, this particular node has both a high-
degree node centrality and a high betweenness centrality,
which suggests that this node is crucial for an unhindered
flow of information in the network. As mentioned previ-
ously, such a node would be a strong candidate for being a
hub in the brain’s large-scale connectome.
We now shift our attention from the simple network
shown in Figure 3A to a full resting-state fMRI data set
acquired in humans. In Figure 3B, an example of the adult
functional brain connectome is shown for a 264-node
parcellation of the brain, with the parcellation shown in the
schematic Figure 2A. Measures of functional brain connec-
tivity were computed based on resting-state fMRI data
acquired from a single subject, namely data from the
Human Connectome 500 subject release (16). Different
resting-state networks are colour-coded using the same
colouring scheme described in the legend to Figure 2 and
shown in Figure 3B. The positioning of the individual brain
nodes in Figure 3B was based on the KamadaKawai
algorithm that produces a so-called spring-embedding
network plot that positions nodes that correlate strongly
with each other nearby in the plot, whereas nodes that have
Figure 3 An illustration of core graph theoretical measures commonly used in brain connectome research. First, the degree for a given node is defined as the sum of all
the edges that are connected to it. In panel A, the node marked with green has a low degree, because it is only one that is connected to one other node. In contrast, the
node marked in orange has a large number of other nodes connected to it and has thus a high degree. Second, the betweenness measure is an estimate of integration
for each node in the network, because a given node is equal to the number of shortest paths from all nodes to all other nodes that pass through that node. Thus, it can
be said that nodes that have a high degree of betweenness centrality possess a large influence on the transfer of information within the network. This is exemplified by
the node depicted in purple, which has both a high degree many edges attached to it but also a high betweenness centrality because a transfer of information
between the upper four nodes and the lower four nodes must pass through this particular node. In panel B, a KamadaKawai plot is shown of the brain connectivity
network of the corresponding brain connectivity matrix shown in Figure 2C. The KamadaKawai algorithm produces a so-called spring-embedding network plot that
positions nodes that correlate strongly with each other nearby, whereas nodes that have a weaker degree of connectivity between them are located more distantly. The
example shown in this figure is based on data provided by the Human Connectome Project, WU-Minn Consortium led by principal investigators David Van Essen and
Kamil Ugurbil (1U54MH091657) and funded by the 16 NIH Institutes and Centers that support the NIH Blueprint for Neuroscience Research and by the McDonnell Center
for Systems Neuroscience at Washington University.
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 5
Mevel and Fransson The functional brain connectome of the child
a weaker degree of connectivity in between them are
located more distantly. By employing the spring-embedding
layout of the connectome, it becomes evident that some
brain regions are clustered together, that is strongly corre-
lated during rest. For example, the visual subnetwork (red),
somatomotor areas (orange) and the subnetworks that
encompass associative cortices, such as the default mode
network indicated by the brown dots, the dorsal and ventral
attention networks (dark green and cyan) and the fronto-
parietal and cingulo-opercular networks (purple and yel-
low), are clustered together (i.e. segregation) in Figure 3B.
At the same time, we find that there are edges in the
connectome that binds together different subnetworks and
that this creates a platform for integration of information
between the different subnetworks. The overall appearance
of the network topology in the example connectome shown
in Figure 3B is in agreement with the notion that functional
segregation and integration coexist in the large-scale human
brain connectome. Although we have used functional
parcellation schemes here that were developed for the
adult brain to illustrate the key concepts in brain connec-
tomics, the same methodological procedures apply to the
young brain. It is noteworthy that while several functional
parcellation schemes of the cortex have been developed for
the adult brain, the availability of functional atlases adapted
to younger brain remains scarce.
The functional connectome of the infant brain
The anatomy and histology of the infant brain has been
extensively studied, and we will briefly touch upon some of
the key events before we turn our attention to its functional
network topology. The main event during the foetal phase is
a gathering of thalamic afferents in the frontal, visual,
auditory and somatosensory cortex. Notably, ingrowth of
the thalamocortical axons in the frontal, somatosensory and
visual cortex takes place during the early preterm phase and
the late preterm phase is characterised by an ingrowth of
callosal and long cortico-cortical pathways in the cortex. At
a cellular level, a rapid growth of dendrites occurs during
the first postnatal months and that synaptogenesis, and
synapse elimination, is heterogeneously distributed over
time in the cerebral cortex. A number of studies provide
further references on the neuroanatomical developments
during prenatal life and infancy (1719).
The first study on the resting-state functional MRI brain
connectivity in the infant brain was published in 2007 (20),
and it used a data-driven approach, namely independent
component analysis (ICA), to show that the brain is
organised into several anatomically widespread large-scale
networks at birth. Interestingly, the results showed that
resting-state networks in the infant brain are well developed
in the primary sensory areas, including the visual, sensori-
motor and auditory systems with bilateral activation pat-
terns. However, resting-state networks that encompass
higher order cortices such as the default mode network
were only partly developed in the infant brain. In particular,
resting-state brain connectivity in the anteriorposterior
direction was almost absent, a finding that resonates well
with MRI diffusion tractography studies of white matter
fibre tracts in infants (21). The overall pattern of structural
brain connectivity in the infant brain includes a rather
strong cross-hemispherical connectivity pattern in the
primary cortices that is accompanied by substantially
weaker network connectivity in higher order cortices. This
network connectivity is gradually strengthened during the
first two years of life, a finding that has been corroborated
by several studies (2224) and commented on in review
papers (2527).
So far, we have discussed the presence of separate
resting-state subnetworks in the infant brain. We now
discuss results that pertain to the infant functional brain
connectome assessed by resting-state fMRI and graph
theory. In the first graph theoretical study of the infant,
the brain connectome was modelled by a connectivity
matrix of size 4.096 by 4.096 nodes, where each node
represented the resting-state fMRI signal from a
voxel residing in grey brain matter (28).
The schematic outline shown in Figure 3 shows that the
BOLD fMRI signal intensity time courses were extracted
from each node, and then, the correlation coefficient was
computed between all pairs of nodes and inserted into a
connectivity matrix (Fig. 2C). The degree of connectivity
between any two nodes was considered significant if its
correlation coefficient exceeded a threshold of 0.3. This
resulted in a binary, unweighted connectivity matrix for
each subject. First, the node degree centrality (Fig. 3A) was
computed for each node (voxel) based on their standardised
scores across subjects. Second, the node betweenness
centrality (Fig. 3A) was computed on a node-by-node basis.
Third, a seed-based correlation analysis was carried out
based on the local maxima in the spatial distribution of the
node degree centrality. The BOLD fMRI signal intensity
time course for the seed was correlated with all other voxels
in the brain.
The spatial distribution of node degree and betweenness
centrality is shown for both infants and adults in Figure 4A
D. As mentioned previously, nodes that score highly on
betweenness centrality and node degree centrality are
considered to be strong candidates for being cortical hubs
in the brain. In infants, candidates for cortical hubs are
primarily located in the primary sensorimotor, auditory and
visual cortices and, to a much lesser extent, in the higher
associative cortex. In contrast, candidates for cortical hubs
in adults are found foremost in the heteromodal cortex,
including the posteromedial parietal cortex, medial pre-
frontal, insular and temporal cortex. The green circles in
Figure 4E,F are examples of the networks associated with
candidates for cortical hubs in infants and adults, respec-
tively. In infants, the seed region positioned in the left
primary somatomotor cortex was significantly correlated
with the lateral and medial aspect of the somatomotor
cortex in both hemispheres. In adults, a seed located in the
posteromedial parietal cortex, a brain region that showed a
high node degree and central betweenness, connected
significantly with other brain regions that are commonly
observed to belong to the brain’s default mode network,
6©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
including the bilateral parietal lobule and medial prefrontal
cortex. Moreover, attempts have been made to delineate the
functional brain connectome of the foetus. Despite the
technical challenges involved, a study could show strong,
general cross-hemispherical functional connectivity. Fur-
thermore, one study showed that an age dependence during
the foetal age was observed for about half of the regions
examined (29). These findings support the notion that the
development of the functional human brain connectome
has already started during foetal age. However, the inter-
pretation of candidate hubs in neurodevelopmental terms is
not straightforward. For example, it is conceivable that the
functional hubs detected at infancy in primarily sensorimo-
tor regions are still acting as local hubs during adulthood,
but that they are being overshadowed by the presumably
much stronger and richer connectivity patterns that have
gradually been established in the higher order cortices
during childhood. So when researchers study neurodevel-
opmental processes, the presence of cortical hubs, in
particular, must be assessed and evaluated in relative terms.
We also need to acknowledge that methodological factors
may play a role when assessing the spatial topology of
functional brain networks. To summarise, both the adult
and infant brain showed a strong presence of cortical hubs,
but their spatial topology was vastly different. Whereas hubs
in adults were primarily found in the higher associative
cortex, including the default mode network and fronto-
parietal networks, hubs in the infant brain were primarily
found in the primary sensory, motor, auditory and visual
brain areas (28).
The development of the functional brain connectome
during childhood
As detailed findings have already been published, this
section provides a brief summary of key results that describe
the typical neurodevelopmental changes in the connectome
during childhood. Collin and van den Heuvel (25) have
provided an exhaustive overview of the biological mecha-
nisms driving the development of the connectome, and
more details on changes in the topological properties of
the developing connectome can be found in V
ertes and
Bullmore’s paper (30). Briefly,it has been suggested that the
variability of large-scale network brain connectivity is under
genetic control, and thus, the heritability is between 40 and
Figure 4 Identification of cortical hubs and their corresponding functional networks in the adult and infant brain, as revealed by resting-state functional magnetic
resonance imaging. The spatial distribution of the magnitude of the degree centrality measure the number of edges connected to each node, defined here as
voxels is shown in panels A (adults) and B (infants). In adults, a high degree of centrality can be observed to reside foremost in the default mode
network regions, including the precuneus/posterior cingulate cortex and the medial prefrontal cortex. In contrast, infants showed a high degree of centrality, which was
most prominent in the primary sensory cortices. The degree of betweenness centrality, which is commonly used as an estimate for putative candidates of cortical hubs, is
shown in panels C (adults) and D (infants). Similar to the spatial distribution of degree centrality, candidate hub regions showed a preference for the default network
regions in the adult brain, whereas primary sensory regions dominated the spatial profile of degree betweenness in infants. Panel E (adults) and F (infants) show typical
examples of the spatial pattern networks associated with candidates for cortical hubs. A seed region (green sphere in the figure) positioned in the precuneus/posterior
cingulate cortex, and a corresponding seed-based correlation analysis, yielded a connectivity pattern that stretched between the brain regions within the default network
regions in adults. Analogously, a seed positioned in the left primary sensorimotor area, a likely candidate for a hub region in infants (panels B and D), produced a robust
connectivity with the bilateral primary sensory network. Figure adapted from Fransson et al. (28).
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 7
Mevel and Fransson The functional brain connectome of the child
60% (31). For instance, the average global path length the
average number of intermediate nodes visited when finding
a path between two given nodes in the connectome was
found to be linked to genes in 42% of monozygotic twins
aged 12 years compared to dizygotic twins (32). This result
is still lower than the degree of heritability for the structural
connectome, of up to 76%, suggesting that environmental
and, or, epigenetic factors, such as education and learning,
might be responsible for nearly 50% of the individual
variability in large-scale brain functional connectivity.
Similarly, candidate genes that are invoked in shaping the
functional architecture of the brain have been identified,
including the MET,COMT,APOE and ZNF804A genes, as
detailed in the review by Thompson et al. (31).
Moreover, previous studies have suggested that the
neurodevelopmental trajectory of brain connectivity
evolves according to an inverted U-shape pattern, with
childhood characterised by an increase in integration
between and within networks. Early adulthood and adult-
hood can best be described by a plateau phase that is
followed by a decrease in network integration at later ages,
that is an increase in local clustering (25). For example, Cao
et al. (33) investigated a large cohort of typically developed
individuals aged seven to 85 years. The authors specifically
investigated the rich club model organisation of the brain,
which postulates that strongly connected nodes in the
connectome show a strong tendency to connect with nodes
that display a similarly high degree of connectedness.
Collectively, the rich club of brain nodes constitutes a
cluster of nodes that serves as a backbone platform for
information transfer in the brain (34). An interesting
property in this context is the so-called efficiency of the
connectome, which can be estimated at a global and local
level. Global efficiency is essentially defined as the inverse
of the average path length between all nodes in the
connectome, and it can be interpreted, in causal terms, as
a measure of how efficient parallel information is trans-
ferred throughout the connectome. At a local level single
nodes the efficiency can be viewed as how well a node
in the connectome tolerates errors and the efficiency of
information transfer among its neighbours when the
affected node is removed. Together with measures of local
efficiency of brain connectivity, the authors showed that the
rich club node organisation at a local level followed an
inverted U-shaped pattern, which peaked at around the age
of 40 years (33). However, it should be noted that the
inverted U-shape relationship between age and network
efficiency does not hold for all scales of network analysis. If
we examine the age dependence on brain network effi-
ciency across the lifespan at the level of individual nodes
and hubs, some nodes actually show the inverted U-shaped
pattern, (e.g. lateral frontal, parietal and temporal regions
and cuneus), while others show either linear decreases
(medial prefrontal cortex, insula, visual cortex and subcor-
tical regions) or increases (linear trend: supplementary
motor and temporal areas and a quadratic trend: parahip-
pocampus, thalamus) in connectivity as a function of age
(33). This finding highlights the complexity of the functional
brain connectome, calling for comprehensive investigations
of its regional topological properties.
Structural and functional connectivity of the brain and its
relationship to learning
While the ontogeny of the structural brain connectome is
largely beyond the scope of this review, and has been
exhaustively reviewed by Tymofieva et al. (27), a highly
relevant issue when describing the development of the
functional connectome is the putative coupling between the
structural and functional connectome. As previously men-
tioned, the brain’s axonal wiring pattern, and the concomi-
tant functional interactions, does not come into existence in
a mature state. While the exact relationship between
structural and functional connectivity in the brain is still
an active area of research, some observations support the
idea that large-scale, resting-state functional connectivity is
partly dependent on the underlying structural connectome
(3538). However, it is very unlikely that a one-to-one
match between anatomy and functional connectivity exists,
because the emergence of functional connectivity builds
upon a combination of several different mechanisms that
each have their own temporal and spatial dynamical
trajectories. For example, it is likely that functional con-
nectivity, as revealed by resting-state fMRI, follows mono-
synaptic as well as poly-synaptic routes for transferring
information in the brain (39). Accordingly, the develop-
mental trajectory of the functional connectome during
childhood and adolescence is expected to differ from the
corresponding path in the structural brain connectome.
Indeed, the specific inverted U-shape pattern mentioned
earlier is assumed to be the result of several biological
mechanisms, including some that have been at work since
prenatal life. Although the findings remain unclear about its
onset, in prenatal or early postnatal life, the increasing
myelination of particularly long-range connections might
support increases in brain networks global efficiency during
the first two years of life. However, there are regional
differences in the timing of maturation that might be related
to developmental changes in the functional connectome
topology during these years. The axonal pathways that
connect the primary cortical regions are among the earliest
to mature, at around three months, while it takes until the
twelfth month of postnatal life for association tracts to
become well delineated. As mentioned in the section on the
infant’s brain, the neonatal functional connectome appears
particularly supportive of primary functions shortly after
birth. In the subsequent two years, developmental changes
in connectome topology support increasingly integrated
information processing, which is paralleled by the develop-
ment of cognitive functions. During later childhood and
adolescence, the brain structural architecture has reached
relative maturity and the main hubs are fully efficient
(35,40,41). This means that the end of the exuberant axon
removal phase allows for adjustments in the connections
strength, that is in the diameter of the axonal bundles, the
relative alignment of individual axons and the density and
myelination. Accordingly, a longitudinal diffusion MRI
8©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
study suggested that only a small selection of brain regions,
approximately 7%, showed increases or decreases in struc-
tural connectivity during adolescence and early adulthood.
On the other hand, this rather small selection of connec-
tions was tied up with 90% of the remaining nodes in the
brain connectome (42). In contrast, segregation (clustering)
has been shown to decrease during childhood and adoles-
cence, while integration (node strength and efficiency)
increased in the functional connectome, resulting in a
topology that is increasingly capable of facilitating higher
order cognitive functions (25). The progressive maturation
of hubs subtending the development of cognitive functions
might also explain differences in the developmental trajec-
tories across graph theoretical measures of the functional
connectome, as rich club organisation only peaks at around
40 years of age in some major hubs (33). Finally, the
possibility that differential influences from environmental
factors on both structural and functional connectivity might
translate into differences in cognitive skills and abilities
later in life must also be considered. A compelling example
of the different developmental trajectories between the
structural and functional connectome has been reported
(40). In this study, they investigated the rich club organi-
sation of both structural and functional nodes in the
connectomes of children aged seven to 11 years, which
were compared to the corresponding nodes in young adults,
aged 2435 years. In the younger cohort, they reported a
significantly lower degree of functional rich club organisa-
tion among connections that linked the superior parietal
lobule, insular and supramarginal cortex. Interestingly, no
group differences could be detected for the structural
connectome (40). This observation suggests that processes
related to functional, but not structural, integration of brain
networks still develop to some degree between 11 and
24 years of age. Moreover, these findings resonate well with
the idea that the central executive or fronto-parietal resting-
state network in typically developed populations maturates
late in life. The central executive network encompasses the
supramarginal and prefrontal cortices, the latter areas being
adjacent to the insular cortex. This network is assumed to
be of major importance, as it underpins executive control,
which is considered to be crucial for the ability to learn. For
example, it actively maintains and manipulates information
in working memory, inhibition of distractive events while
reasoning and decision making (43,44). The prefrontal
cortex has long been considered to be one of the hallmarks
of the later evolutionary changes that have taken place in
the human brain, as discussed by Goulas et al. (45). The
integration of the fronto-parietal network mostly relies on
the maturation of the prefrontal cortex that occurs the latest
in life (46,47). Hence, it is not overly surprising that the
findings from connectomics research that have been
reported so far point towards the fact that a functional
integration of the prefrontal cortices occur at a later age
(40). Additional information regarding changes in the
functional brain connectome during childhood through
adolescence can be found in three studies (4850).
Investigations of putative atypical connectome patterns
in ASD
Disturbances of brain connectivity can occur at any phase
in life, from the early neurodevelopmental disorders to age-
induced neurodegenerative diseases. As demonstrated in
the previous examples, the connectome framework is a
powerful and versatile tool that explores both the spatial
and temporal deviations of the topology of the functional
brain connectome that may be introduced by atypical
neurodevelopmental processes. Graph theoretical explo-
rations of the brain networks have therefore been applied to
several developmental disorders, including attention-deficit
hyperactivity disorder (ADHD) (51) and schizophrenia
(52), in an attempt to separate individuals with disorders
from typically developed children. A good example of how
the connectome concept might help to refine our knowl-
edge about neurodevelopmental disorders relates to the
hypo-/hyperconnectivity hypothesis in ASD. Again, the aim
of the present review was not to carry out an exhaustive
review of the literature, but to give examples on how the
connectome framework could be advantageously used to
gain an improved understanding of the neurobiological
underpinnings of ASD. ASD is a highly prevalent neuro-
developmental condition, which occurs in around 12% of
school-aged children (53,54), and it is defined by disabling
impairments in social communication and interaction that
occur alongside restricted interests and repetitive beha-
viours, according to the American Psychiatric Association
(55). As the phenotype, biology and aetiology of ASD are
heterogeneous, very little is known about the neuronal
mechanisms and developmental processes that might be
the driving forces for the behavioural patterns typically
observed in ASD. In an attempt to further elucidate putative
neurobiological mechanisms, Just et al. (56) proposed the
hypoconnectivity theory of ASD. According to their
hypothesis, ASD might be associated with an overall
under-functioning of the brain’s integrative circuitry that
results in a deficit of integration of neuronal information at
the neural level. Subsequently, Courchesne and Pierce (57)
suggested an alternative hypothesis that postulated that the
developmental trajectories for functional brain connectivity
in ASD individuals was characterised by both an early local
hyperconnectivity and a long-distance hypoconnectivity of
the prefrontal cortex. That is an increased short-range
(local) connectivity within the frontal lobe but a decreased
degree of functional long-range (global) connectivity with
the other parts of the brain. Courchesne et al. further
proposed that their hypothesis, if true, suggested that the
connectivity pattern in the frontal lobe in ASD individuals
could be characterised as being excessive, disorganised and
inadequately selective. In contrast, connectivity between
the frontal cortex and other systems is assumed to be poorly
synchronised and weakly responsive, leading to poor
information flow. Subsequent investigations of the connec-
tivity patterns in ASD individuals focused on the presumed
dichotomy of brain organisation, seeking to find further
support for such a dichotomy in the early stages of
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 9
Mevel and Fransson The functional brain connectome of the child
neurodevelopment. Despite the rather extensive literature
published so far on ASD in children and adolescents, a
coherent picture on how brain connectivity is affected in
ASD is still lacking. Although there is some support for the
idea of a switch from a global hyperconnectivity to a
hypoconnectivity pattern in ASD individuals at around the
age of 12 years (58), both local and global functional
profiles have been reported in young ASD populations.
Detailed reviews on brain connectivity patterns in ASD
populations have been published (5860). It should be
noted that several methodological confounds and limita-
tions might have influenced the partly conflicting results,
including differences in the sample size or composition,
inadequate corrections for micro-head movements and,
importantly, a lack of control for age range and a scarcity of
data acquired in longitudinal sample cohorts. According
to the review paper by Uddin et al. (58), discrepancies
between findings of autism-related hypoconnectivity and
hyperconnectivity might be reconciled by taking develop-
mental changes, such as puberty, into account. Based on
results about the impact of hormones onto the brain
architecture and functioning, they proposed that longitu-
dinal studies examining prepubertal and postpubertal
individuals were much needed to resolve the current
inconsistencies regarding the true nature of abnormal
development of large-scale functional brain connectivity
in ASD. For the reasons described above, we chose to focus
upon ASD in the next section. In the absence of investiga-
tions relying on prepubertal and postpubertal connectivity
differences in ASD individuals as described earlier, all these
call for further innovative and comprehensive explorations
of the neurodevelopmental trajectory associated with ASD,
which the functional connectome might offer.
Novel insights of the neurobiology behind ASD gained
from brain connectome research
A functional brain connectome that is sufficiently accurate
and representative of ASD has not been fully mapped yet,
and there is speculation about its biologically relevant
characteristics. However, preliminary findings remain
promising. It has been suggested that the ASD connectome
is characterised by a combination of both ectopic and
immature connectivity (26). It has also been suggested that
ASD is the consequence of abnormal brain development,
rather than as an abrupt alteration of a normally formed
network (S61). Consequently, the analysis of the so-called
miswired ASD brain connectome (26) would need to
involve both a description of the diagnostic abnormalities
at the level of the connectome and an understanding of
how these abnormalities could have been produced as an
aberrant expression of the normal processes of brain
network development. In sharp contrast to their hetero-
geneity, the classical atypical functional connectivity pat-
terns reported in ASD have mostly been shown to be
associated with symptom severity (59,60). Accordingly, the
association between the functional measures of the brain
connectome and the severity of symptoms also appears to
be a promising lead. Unfortunately, the results that have
been reported so far remain unclear. For instance, Redcay
et al. showed that the parietal nodes of the default mode
network had higher betweenness centrality measures in
adolescents and young adults with ASD, while their
connectivity with anterior nodes was associated with
poorer communication skills (S62).Reversed patterns were
also reported, with ASD children and adolescents showing
reduced functional integration of the default mode network
without any direct correlation to the severity of their
symptoms (S63). In parallel, novel findings might also be
gained from prospective explorations of brain connectivity
in infants at risk for ASD, as such having older sibling
diagnosed with ASD. A retrospective comparison of con-
nectivity profiles during infancy to the functional patterns
observed at the time of clinical diagnosis, for example such
as by the age of three years, might provide an opportunity to
identify biomarkers that appears before the onset of clinical
symptoms. To the best of our knowledge, such a prospective
design remains to be tested in combination with connec-
tome and graph theoretical modelling of brain functional
It seems unlikely that the methodological tools, as well as
the quantity of neuroimaging data that are currently
available, will be sufficient to achieve the acceptable degree
of accuracy we need to understand neurodevelopmental
disorders from a brain connectivity perspective. However,
we believe that there are several good reasons why we
should remain optimistic regarding what future research on
the child brain connectome might contribute to our capa-
bility to scrutinise the large-scale neuronal processes that
shapes the developing human brain. First, we should
mention that research on large-scale brain connectomics
is still relatively new. Strategies for data preprocessing, data
analysis and graph theoretical modelling of neuroimaging
data are continuously being refined and optimised to
investigate neurodevelopmental processes. An example of
methodological work to address these issues is the research
that aims to achieve higher accuracy for handling non-
neuronal signal sources, such as subject head movement
(S64) and other potential sources of signal bias in fMRI
With regard to methodological developments, we would
like to briefly mention efforts to investigate the temporal
dynamics of functional brain connectivity. Static functional
connectivity, which is the parameter of brain connectivity
that we have concerned ourselves with so far, is based upon
computing a single estimate of brain connectivity, that is the
correlation coefficient, computed on the entire BOLD fMRI
signal intensity time series. In contrast, dynamic functional
brain connectivity aims to measure temporal fluctuations in
the degree of connectivity and to subsequently use cluster-
ing techniques to classify dynamic changes in connectivity
into different states of brain activity (S65). The pursuit of
tracking dynamic changes in brain connectivity has gained
a considerable degree of interest in a number of studies
10 ©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
(S66S69), and its applicability to neurodevelopmental
questions has been suggested (S70). But perhaps the largest
boost to research on the child brain connectome will come
from the analysis of very large structural and functional
neuroimaging data sets of up to several thousand individ-
uals that have been made publically available to research-
ers. New data are continually being added. These include
studies by Kennedy et al. and Satterthwaite et al. (S71,S72).
We think that the study of the complex relationship
between genetics, behavioural traits, structural and func-
tional child brain connectivity will substantially benefit
from the analysis of large multidimensional data sets.
None. The funders mentioned in Figure 3 had no role in
study design, data collection and analysis, decision to
publish, or preparation of the manuscript.
PF was supported by a grant from the Swedish Research
1. Sporns O, Tononi G, K
otter R. The human connectome: a
structural description of the human brain. PLoS Comput Biol
2005; 1: 24551.
2. Van Dijk KRA, Hedden T, Venkataraman A, Evans KC, Lazar
SW, Buckner RL. Intrinsic functional connectivity as a tool for
human connectomics: theory, properties, and optimization.
J Neurophysiol 2010; 103: 297321.
3. Damoiseaux JS, Rombouts SA, Barkhof F, Scheltens P, Stam
CJ, Smith SM, et al. Consistent resting-state networks across
healthy subjects. Proc Natl Acad Sci USA 2006; 103: 1384853.
4. Fransson P. How default is the default mode of brain function?
Further evidence from intrinsic BOLD signal fluctuations.
Neuropsychologia 2006; 44: 283645.
5. Smith SM, Fox PT, Miller KL, Glahn DC, Fox PM, Mackey CE,
et al. Correspondence of the brain’s functional architecture
during activation and rest. Proc Natl Acad Sci USA 2009; 106:
6. Biswal BB, Yetkin FZ, Haughton VM, Hyde JS. Functional
connectivity in the motor cortex of resting human brain using
echo-planar MRI. Magn Reson Med 1995; 34: 53741.
7. Leopold DA, Murayama Y, Logothetis NK. Very slow activity
fluctuations in monkey visual cortex: implications for
functional brain imaging. Cereb Cortex 2003; 13: 42233.
8. Shmuel A, Leopold DA. Neuronal correlates of spontaneous
fluctuations in fMRI signals in monkey visual cortex:
implications for functional connectivity at rest. Hum Brain
Mapp 2008; 29: 75161.
9. Sch
olvinck ML, Maier A, Ye FQ, Duyn JH, Leopold DA.
Neural basis of global resting-state fMRI activity. Proc Natl
Acad Sci USA 2010; 107: 1023843.
10. Chang C, Liu Z, Chen MC, Liu X, Duyn JH. EEG correlates of
time-varying BOLD functional connectivity. NeuroImage
2013; 72: 22736.
11. Fornito A, Zalesky A, Breakspear M. The connectomics of
brain disorders. Nat Rev Neurosci 2015; 16: 15970.
12. Zhang D, Raichle ME. Disease and the brain’s dark energy. Nat
Rev Neurol 2010; 6: 1528.
13. Power JD, Cohen AL, Nelson SM, Wig GS, Barnes KA, Church
JA, et al. Functional network organization of the human brain.
Neuron 2011; 72: 66578.
14. Rubinov M, Sporns O. Complex network measures of brain
connectivity: uses and interpretations. NeuroImage 2010; 52:
15. Fox PT, Friston KJ. Distributed processing: distributed
functions? NeuroImage 2011; 61: 40726.
16. Van Essen DC, Smith SM, Barch DM, Behrens TEJ, Yacoub E,
Ugurbil K, et al. The WU-Minn Human Connectome Project:
an overview. NeuroImage 2013; 80: 6279.
17. Huttenlocher PR, Dabholkar AS. Regional differences in
synaptogenesis in human cerebral cortex. J Comp Neurol 1997;
387: 16778.
18. Kostovic I, Jovanov-Milosevic N. The development of cerebral
connections during the first 20-45 weeks’ gestation. Semin
Fetal Neonatal Med 2006; 2006: 41522.
19. Petanjek Z, Judas M, Kostovic I, Uylings HBM. Lifespan
alterations of basal dendritic trees of pyramidal neurons in the
human prefrontal cortex: a layer-specific pattern. Cereb Cortex
2008; 18: 91529.
20. Fransson P, Ski
old B, Horsch S, Nordell A, Blennow M,
Lagercrantz H, et al. Resting-state networks in the infant brain.
Proc Natl Acad Sci USA 2007; 104: 155316.
21. Dubois J, Dehaene-Lambertz G, Kulikova S, Poupon C, Huppi
PS, Hertz-Pannier L. The early development of brain white
matter: a review of imaging studies in fetuses, newborns and
infants. Neuroscience 2014; 276: 4871.
22. Gao W, Zhu H, Giovanello KS, Smith JK, Shen D, Gilmore JH,
et al. Evidence on the emergence of the brain’s default network
from 2-week-old to 2-year-old healthy pediatric subject. Proc
Natl Acad Sci USA 2009; 106: 67905.
23. Smyser CD, Inder TE, Shimony JS, Hill JE, Degnan AJ, Snyder
AZ, et al. Longitudinal analysis of neural network
development in preterm infants. Cereb Cortex 2010; 20: 2852
24. Doria V, Backmann CF, Arichi T, Merchant N, Groppo M,
Turkheimer FE, et al. Emergence of resting-state networks in
the preterm human brain. Proc Natl Acad Sci USA 2010; 111:
25. Collin G, van den Heuvel MP. The ontogeny of the human
connectome: development and dynamic changes of brain
connectivity across the life span. Neuroscientist 2013; 19: 616
26. Di Martino A, Fair DA, Kelly C, Satterthwaite TD, Castellanos
XF, Thomason ME, et al. Unraveling the miswired
connectome: a developmental perspective. Neuron 2014; 83:
27. Tymofiyeva O, Hess CP, Xu D, Barkovich AJ. Structural MRI
connectome in development: challenges of the changing brain.
Br J Radiol 1039; 2014: 20140086. doi:10.1259/bjr.20140086.
28. Fransson P,
Aden U, Blennow M, Lagercrantz H. The
functional architecture of the infant brain as revealed by
resting-state fMRI. Cereb Cortex 2011; 21: 14554.
29. Thomason ME, Dassanayake MT, Shen S, Katkuri Y, Alexis M,
Anderson AL, et al. Cross-hemispheric functional connectivity
in the human fetal brain. Sci Transl Med 2013; 5: 173ra24.
30. V
ertes PE, Bullmore ET. Annual research review: growth
connectomicsthe organization and reorganization of brain
networks during normal and abnormal development. J Child
Psychol Psychiatry 2015; 56: 299320.
31. Thompson PM, Ge T, Glahn DC, Jahanshad N, Nichols TE.
Genetics of the connectome. NeuroImage 2013; 80: 47588.
©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd 11
Mevel and Fransson The functional brain connectome of the child
32. van den Heuvel MP, van Soelen ILC, Stam CJ, Kahn RS,
Boomsma DI, Hulshoff Pol HE. Genetic control of
functional brain network efficiency in children. Eur
Neuropsychopharmacol 2013; 23: 1923.
33. Cao M, Wang JH, Dai ZJ, Cao XY, Jiang LL, Fan FM, et al.
Topological organization of the human brain functional
connectome across the lifespan. Dev Cogn Neurosci 2014; 7:
34. van den Heuvel MP, Sporns O. Rich-club organization of the
human connectome. J Neurosci 2011; 31: 1577586.
35. Hagmann P, Sporns O, Madan N, Cammoun L, Pienaar R,
Wedeen VJ, et al. White matter maturation reshapes structural
connectivity in the late developing human brain. Proc Natl
Acad Sci USA 2010; 107: 1906772.
36. Honey CJ, Sporns O, Cammoun L, Gigandet X, Thiran JP,
Meuli R, et al. Predicting human resting-state functional
connectivity from structural connectivity. Proc Natl Acad Sci
USA 2009; 106: 203540.
37. Skudlarski P, Jagannathan K, Calhoun VD, Hampson M,
Skudlarska B, Pearlson G. Measuring brain connectivity:
diffusion tensor imaging validates resting state temporal
correlations. NeuroImage 2008; 43: 55461.
38. van den Heuvel MP, Mandl RCW, Kahn RS. Hulshoff Pol HE.
Functionally linked resting-state networks reflect the
underlying structural connectivity architecture of the human
brain. Hum Brain Mapp 2009; 30: 312741.
39. Sporns O. Structure and function of complex brain networks.
Dialogues Clin Neurosci 2013; 15: 24762.
40. Grayson DS, Ray S, Carpenter S, Iyer S, Dias TGC, Stevens C,
et al. Structural and functional rich club organization of the
brain in children and adults. PLoS ONE 2014; 9: e88297.
41. Hwang K, Hallquist MN, Luna B. The development of hub
architecture in the human functional brain network. Cereb
Cortex 2013; 23: 238093.
42. Baker ST, Lubman DI, Yucel M, Allen NB, Whittle S, Fulcher
BD, et al. Developmental changes in brain network hub
connectivity in late adolescence. J Neurosci 2015; 35: 907887.
43. Houd
e O, Rossi S, Lubin A, Joliot M. Mapping numerical
processing, reading, and executive functions in the developing
brain: an fMRI meta-analysis of 52 studies including 842
children. Dev Sci 2010; 13: 87685.
44. Uddin L, Supekar K, Ryali S, Menon V. Dynamic
reconfiguration of structural and functional connectivity across
core neurocognitive brain networks with development.
J Neurosci 2011; 31: 1857889.
45. Goulas A, Bastiani M, Bezgin G, Uylings HBM, Roebroeck A,
Stiers P. Comparative analysis of the macroscale structural
connectivity in the macaque and human brain. PLoS Comp
Biol 2014; 10: e1003529.
46. Casey BJ, Tottenham N, Liston C, Durston S. Imaging the
developing brain: what have we learned about cognitive
development? Trends Cogn Sci 2005; 9: 10410.
47. Gogtay N, Giedd JN, Lusk L, Hayashi KM, Greenstein D,
Vaituzis C, et al. Dynamic mapping of human cortical
development during childhood through early adulthood. Proc
Natl Acad Sci USA 2004; 101: 81749.
48. Vogel AC, Power JD, Petersen SE, Schlaggar BL. Development
of the brain’s functional network architecture. Neuropsychol
Rev 2010; 20: 36275.
49. Power JD, Fair DA, Schlaggar BL, Petersen SE. The
development of human functional brain networks. Neuron
2010; 67: 73548.
50. Uddin LQ, Supekar K, Menon V. Typical and atypical
development of functional human brain networks: insights
from resting-state FMRI. Front Syst Neurosci 2010; 21: 421.
51. Matthews M, Fair DA. Research Review: functional brain
connectivity and child psychopathology Overview and
methodological considerations for investigators new to the
field. J Child Psychol Psychiatry 2015; 56: 40014.
52. Schmidt A, Diwadkar VA, Smieskova R, Harrisberger F, Lang
UE, McGuire P, et al. Approaching a network connectivity-
driven classification of the psychosis continuum: a selective
review and suggestions for future research. Front Hum
Neurosci 2014; 8: 1047.
53. Idring S, Rai D, Dal H, Dalman C, Sturm H, Zander E, et al.
Autism spectrum disorders in the Stockholm Youth Cohort:
design, prevalence and validity. PLoS ONE 2012; 7: e41280.
54. Kim YS, Leventhal BL, Koh YJ, Fombonne E, Laska E, Lim
EC, et al. Prevalence of autism spectrum disorders in a total
population sample. Am J Psychiatry 2011; 168: 90412.
55. American Psychiatric Association (APA). Diagnostic and
statistical manual of mental disorders. 5th ed. Washington,
DC: American Psychiatric Association, 2013
56. Just MA, Cherkassky VL, Keller T, Minshew NJ. Cortical
activation and synchronization during sentence
comprehension in high-functioning autism: evidence of
underconnectivity. Brain 2004; 127: 181121.
57. Courchesne E, Pierce K. Why the frontal cortex in autism might
be talking only to itself: local over-connectivity but long-
distance disconnection. Curr Opin Neurobiol 2005; 15: 22530.
58. Uddin LQ, Supekar K, Menon V. Reconceptualizing functional
brain connectivity in autism from a developmental perspective.
Front Hum Neurosci 2013; 7: 458.
59. Mevel K, Fransson P, B
olte S. Multimodal brain imaging in
autism spectrum disorder and the promise of twin research.
Autism 2015; 11: 21008.
60. Vissers ME, Cohen MX, Geurts HM. Brain connectivity and
high functioning autism: a promising path of research that
needs refined models, methodological convergence, and
stronger behavioral links. Neurosci Biobehav Rev 2012; 36:
Additional Supporting Information may be found in the
online version of this article:
Data S1 Supplementary references.
12 ©2016 Foundation Acta Pædiatrica. Published by John Wiley & Sons Ltd
The functional brain connectome of the child Mevel and Fransson
... response inhibition, mental flexibility, sustained attention and working memory) (Demetriou et al. 2018;Karalunas et al. 2018;Thamotharan et al. 2013;Wu et al. 2017). Corroborating these functional deficits, functional neuroimaging experiments have suggested that each of ASD and OWOB seems to be associated with abnormalities in resting state functional magnetic resonance imaging (rs-fMRI) connectivity in two neuronal networks involved in regulation of social cognition and executive control, known as the default-mode network (DMN) and the central executive network (CEN) respectively (Hull et al. 2016;Mevel and Fransson 2016;Uddin et al. 2010). Specifically, hypoconnectivity was observed between the anterior (i.e. ...
... Brain rs-fMRI functional connectivity, especially in the DMN and CEN, provides strong neurophysiological and mechanistic evidence of regulation of social cognition and executive control respectively (Hull et al. 2016;Mevel and Fransson 2016;Uddin et al. 2010). As described above, brain rs-fMRI functional connectivity of DMN and CEN in isolated ASD and OWOB has been established. ...
Full-text available
Evidence on neurophysiological correlates of coexisting autism spectrum disorders (ASD) and overweight/obesity may elucidate mechanisms leading to the observed greater risk of obesity in children with ASD. An exploratory secondary data analysis was performed on resting state functional magnetic resonance imaging (rs-fMRI) data of children downloaded from the ABIDE Preprocessed database (n = 81). Children with isolated ASD showed hypo-connectivity between anterior and posterior default mode network (DMN) (p = 0.003; FWER). Children with coexisting ASD and overweight/obesity showed hyper-connectivity between anterior and posterior DMN (p = 0.015; FWER). More evidence is needed to confirm these contrasting rs-fMRI connectivity profiles and to explicate causal inferences regarding neurophysiological mechanisms associated with coexisting ASD and overweight/obesity.
... The connectome model, although still current and whose validity has been confirmed by studies that found similar evidence in terms of altered connectivity, could be considered "encephalocentric" (97,98). On the contrary, according to the "connectivome theory, " elements of a somatic nature that were not adequately valued in the connectome model now take on an equally important role, without the hierarchization of soma and psyche. ...
Full-text available
The classical approach to autism spectrum disorders (ASD) is often limited to considering their neuro-functional aspects. However, recent scientific literature has shown that ASDs also affect many body systems and apparatuses such as the immune system, the sensory-motor system, and the gut-brain axis. The connective tissue, a common thread linking all these structures, may have a pathogenetic role in the multisystem involvement of ASD. Depending on its different anatomical sites, the connective tissue performs functions of connection and support; furthermore, it acts as a barrier between the external and internal environments, regulating the interchange between the two and performing immunological surveillance. The connective tissue shares a close relationship with the central nervous system, the musculoskeletal system and the immune system. Alterations in brain connectivity are common to various developmental disorders, including ASD, and for this reason here we put forward the hypothesis that alterations in the physiological activity of microglia could be implicated in the pathogenesis of ASD. Also, muscle hypotonia is likely to clinically correlate with an altered sensoriality and, in fact, discomfort or early muscle fatigue are often reported in ASDs. Furthermore, patients with ASD often suffer from intestinal dysfunctions, malabsorption and leaky gut syndrome, all phenomena that may be linked to reduced intestinal connectivity. In addition, at the cutaneous and subcutaneous levels, ASDs show a greater predisposition to inflammatory events due to the lack of adequate release of anti-inflammatory mediators. Alveolar-capillary dysfunctions have also been observed in ASD, most frequently interstitial inflammations, immune-mediated forms of allergic asthma, and bronchial hyper-reactivity. Therefore, in autism, altered connectivity can result in phenomena of altered sensitivity to environmental stimuli. The following interpretative model, that we define as the “connectivome theory,” considers the alterations in connective elements of common mesodermal origin located in the various organs and apparatuses and entails the evaluation and interpretation of ASDs through also highlighting somatic elements. We believe that this broader approach could be helpful for a more accurate analysis, as it is able to enrich clinical evaluation and define more multidisciplinary and personalized interventions.
... In general, network development tends to follow a localto-global pattern whereby distant regions and networks become more connected with age (Fair et al. 2009;Power et al. 2010;Vogel et al. 2010). The decreased connectivity between networks seen here suggests that children and adolescents with PAE do not show the expected interactions among brain regions (Fair et al. 2012;Rudie et al. 2013;Matthews and Fair 2015;Mevel and Fransson 2016). Previous PAE studies examining individual networks have described reduced functional connectivity in the default mode network (Santhanam et al. 2011), regions within the frontal parietal network and the salience network , and somatomotor networks (Donald et al. 2016;Long et al. 2018). ...
Full-text available
Diffuson tensor imaging (DTI) has demonstrated widespread alterations of brain white matter structure in children with prenatal alcohol exposure (PAE), yet it remains unclear how these alterations affect the structural brain network as a whole. The present study aimed to examine changes in the DTI-based structural connectome in children and adolescents with PAE compared to unexposed controls. Participants were 121 children and adolescents with PAE (51 females) and 119 typically-developing controls (49 females) aged 5–18 years with DTI data collected at one of four research centers across Canada. Graph-theory based analysis was performed on the connectivity matrix constructed from whole-brain white matter fibers via deterministic tractography. The PAE group had significantly decreased whole-brain global efficiency, degree centrality, and participation coefficients, as well as increased shortest path length and betweenness centrality compared to unexposed controls. Individuals with PAE had decreased connectivity between the attention, somatomotor, and default mode networks compared to controls. This study demonstrates decreased structural white matter connectivity in children and adolescents with PAE at a whole-brain level, suggesting widespread alterations in how networks are connected with each other. This decreased connectivity may underlie cognitive and behavioural difficulties in children with PAE.
... Firstly, previous studies have suggested that neurobiological dysfunction was believed to be important in ASD. Previous studies have also suggested the abnormal acceleration of brain growth in early childhood with ASD (15), and disorders in consistent networks of brain regions, especially the social network, were closely linked with ASD (31)(32)(33). As highlighted in an increasing number of functional MRI (fMRI) articles, the atypical connectivity of brain networks in ASD included the superior temporal gyri (34), inferior frontal gyri (34), insula (35), default mode network (36), cingulate region (37), and parieto-occipital region (38). ...
Full-text available
Autism spectrum disorder (ASD) is a neurodevelopmental disease that may involve various brain abnormalities. However, there are few large epidemiological studies on the relation between epilepsy and ASD in terms of different genders and ages. This study aimed to evaluate the relation between epilepsy and ASD based on 74,251 Chinese children aged 3–12 years who were recruited from kindergartens and primary schools in China. ASD was diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders—Fifth Edition (DSM-V), and verification of epilepsy was based on medical records. The enrolled children diagnosed with ASD were examined by magnetic resonance imaging (MRI) and took genetic tests to rule out other neurological and congenital diseases. The raw odds ratio (OR) was 60.53 [95% confidence interval (CI) = 37.80–96.92, P < 0.01] for epilepsy and ASD, and the adjusted OR was 38.99 (95% CI = 20.70–73.41, P < 0.01) after controlling for the confounders. Moreover, the adjusted OR was significantly higher in girls (OR = 45.26, 95% CI = 16.42–124.76, P < 0.01) than in boys (OR = 32.64, 95% CI = 14.33–74.34, P < 0.01). Among children with younger age, the adjusted OR was the highest (OR = 75.12, 95% CI = 22.80–247.48.16, P < 0.01). These findings suggest that epilepsy might be closely linked to the development of ASD, especially for early-onset epilepsy and among girls.
... The current findings point out the limits of modular approaches to understanding regional specificity in ASD. As improvement in connectomic research and ASD progress (see Sathyanesan et al. 2019;Mevel and Fransson 2016), an improved understanding of which brain areas contribute to VMI performance and impairment will likely emerge. ...
Although diminished proficiency on tasks that require visual-motor integration (VMI) has been reported in individuals with autism spectrum disorder (ASD), very few studies have examined the association between VMI performance and neuroanatomical regions of interest (ROI) involved in motor and perceptual functioning. To address these issues, the current study included an all-male sample of 41 ASD (ages 3–23 years) and 27 typically developing (TD) participants (ages 5–26 years) who completed the Beery-Buktenica Developmental Test of Visual-Motor Integration (Beery VMI) as part of a comprehensive neuropsychological battery. All participants underwent 3.0 T magnetic resonance imaging (MRI) with image quantification (FreeSurfer software v5.3). The groups were statistically matched on age, handedness, and intracranial volume (ICV). ASD participants performed significantly lower on VMI and IQ measures compared with the TD group. VMI performance was significantly correlated with FSIQ and PIQ in the TD group only. No pre-defined neuroanatomical ROIs were significantly different between groups. Significant correlations were observed in the TD group between VMI and total precentral gyrus gray matter volume (r = .51, p = .006) and total frontal lobe gray matter volume (r = .46, p = .017). There were no significant ROI correlations with Beery VMI performance in ASD participants. At the group level, despite ASD participants exhibiting reduced visuomotor abilities, no systematic relation with motor or sensory-perceptual ROIs was observed. In the TD group, results were consistent with the putative role of the precentral gyrus in motor control along with frontal involvement in planning, organization, and execution monitoring, all essential for VMI performance. Given that similar associations between VMI and ROIs were not observed in those with ASD, neurodevelopment in ASD group participants may not follow homogenous patterns making correlations in these brain regions unlikely to be observed.
... For example, studies have reported early developmental connectome changes in autism spectrum disorder (ASD) 66 , including lower white matter integrity, larger fibre count and increased network path length 24 . Network studies have further reported under-functioning of the brain's integrative circuitry 175,176 , combined with functional over-connectivity of local circuitry in ASD 67 , as well as inter-individual variation in levels of functional connectivity that potentially relates to traits associated with the disorder 19 . For schizophrenia, brain connectivity studies have reported widespread changes in anatomical connectivity [177][178][179] , including alterations in rich club and core architecture in people with the disorder 68,70,179 and their siblings 180 and in individuals at clinically high risk of developing psychosis 143,144 . ...
Many human brain disorders are associated with characteristic alterations in the structural and functional connectivity of the brain. In this article, we explore how commonalities and differences in connectome alterations can reveal relationships across disorders. We survey recent literature on connectivity changes in neurological and psychiatric disorders in the context of key organizational principles of the human connectome and observe that several disturbances to network properties of the human brain have a common role in a wide range of brain disorders and point towards potentially shared network mechanisms underpinning disorders. We hypothesize that the distinct dimensions along which connectome networks are organized (for example, ‘modularity’ and ‘integration’) provide a general coordinate system that allows description and categorization of relationships between seemingly disparate disorders. We outline a cross-disorder ‘connectome landscape of dysconnectivity’ along these principal dimensions of network organization that may place shared connectome alterations between brain disorders in a common framework. In this Opinion article, Martijn van den Heuvel and Olaf Sporns examine alterations in structural and functional brain connectivity across brain disorders. They propose a common landscape for such alterations that is based on principles of network organization.
Functional connectomes, as measured with functional magnetic resonance imaging (fMRI), are highly individualized, and evidence suggests this individualization may increase across childhood. A connectome can become more individualized either by increasing self-stability or decreasing between-subject-similarity. Here we used a longitudinal early childhood dataset to investigate age associations with connectome self-stability, between-subject-similarity, and developmental individualization, defined as an individual's self-stability across a 12-month interval relative to their between-subject-similarity. fMRI data were collected during an 18-minute passive viewing scan from 73 typically developing children aged 4-7 years, at baseline and 12-month follow-up. We found that young children had highly individualized connectomes, with sufficient self-stability across 12-months for 98% identification accuracy. Linear models showed a significant relationship between age and developmental individualization across the whole brain and in most networks. This association appeared to be largely driven by an increase in self-stability with age, with only weak evidence for relationships between age and similarity across participants. Together our findings suggest that children's connectomes become more individualized across early childhood, and that this effect is driven by increasing self-stability rather than decreasing between-subject-similarity.
Full-text available
Autism spectrum disorders (ASDs) are a group of heterogeneous neu-rodevelopmental conditions characterized by deficits in social communication and social interaction and restricted, repetitive patterns of behaviors, interests, or activities. For many years, psychoanalysis and neurobiology have been in opposite camps regarding the understanding of autism in terms of causation and treatment. This paper aims to highlight converging points between neurobiological and psychodynamic understanding of autism, which could be useful in designing more effective early interventions. For this purpose, we give a brief overview of the psychoanalytic conceptualization of autism since its first description as well as present the most pertinent neurobiological findings underlying the disorder; both these approaches are pointing to a dysfunction in caregiver-child interactions. In the last few decades, the convergence of the psychoanalytical with the neurobiological perspectives of the disorder enhances further our understanding of the dynamic interplay among biological and psychological processes in autism. This integrative approach, grounded in both theoretical perspectives, could inform future research focusing on interpersonal neurobiology, but also provide a base for developing multi-level and multi-component early interventions, which should start as early as possible, most appropriately during infancy.
Full-text available
The human brain undergoes substantial development throughout adolescence and into early adulthood. This maturational process is thought to include the refinement of connectivity between putative connectivity hub regions of the brain, which collectively form a dense core that enhances the functional integration of anatomically distributed, and functionally specialized, neural systems. Here, we used longitudinal diffusion magnetic resonance imaging to characterize changes in connectivity between 80 cortical and subcortical anatomical regions over a 2 year period in 31 adolescents between the ages of 15 and 19 years. Connectome-wide analysis indicated that only a small subset of connections showed evidence of statistically significant developmental change over the study period, with 8% and 6% of connections demonstrating decreased and increased structural connectivity, respectively. Nonetheless, these connections linked 93% and 90% of the 80 regions, respectively, pointing to a selective, yet anatomically distributed pattern of developmental changes that involves most of the brain. Hub regions showed a distinct tendency to be highly connected to each other, indicating robust "rich-club" organization. Moreover, connectivity between hubs was disproportionately influenced by development, such that connectivity between subcortical hubs decreased over time, whereas frontal-subcortical and frontal-parietal hub-hub connectivity increased over time. These findings suggest that late adolescence is characterized by selective, yet significant remodeling of hub-hub connectivity, with the topological organization of hubs shifting emphasis from subcortical hubs in favor of an increasingly prominent role for frontal hub regions. Copyright © 2015 the authors 0270-6474/15/359078-10$15.00/0.
Full-text available
Brain changes in schizophrenia evolve along a dynamic trajectory, emerging before disease onset and proceeding with ongoing illness. Recent investigations have focused attention on functional brain interactions, with experimental imaging studies supporting the disconnection hypothesis of schizophrenia. These studies have revealed a broad spectrum of abnormalities in brain connectivity in patients, particularly for connections integrating the frontal cortex. A critical point is that brain connectivity abnormalities, including altered resting state connectivity within the fronto-parietal (FP) network, are already observed in non-help-seeking individuals with psychotic-like experiences. If we consider psychosis as a continuum, with individuals with psychotic-like experiences at the lower and psychotic patients at the upper ends, individuals with psychotic-like experiences represent a key population for investigating the validity of putative biomarkers underlying the onset of psychosis. This paper selectively addresses the role played by FP connectivity in the psychosis continuum, which includes patients with chronic psychosis, early psychosis, clinical high risk, genetic high risk, as well as the general population with psychotic experiences. We first discuss structural connectivity changes among the FP pathway in each domain in the psychosis continuum. This may provide a basis for us to gain an understanding of the subsequent changes in functional FP connectivity. We further indicate that abnormal FP connectivity may arise from glutamatergic disturbances of this pathway, in particular from abnormal NMDA receptor-mediated plasticity. In the second part of this paper we propose some concepts for further research on the use of network connectivity in the classification of the psychosis continuum. These concepts are consistent with recent efforts to enhance the role of data in driving the diagnosis of psychiatric spectrum diseases.
Full-text available
We first give a brief introduction to graph theoretical analysis and its application to the study of brain network topology or connectomics. Within this framework, we review the existing empirical data on developmental changes in brain network organization across a range of experimental modalities (including structural and functional MRI, diffusion tensor imaging, magnetoencephalography and electroencephalography in humans). We discuss preliminary evidence and current hypotheses for how the emergence of network properties correlates with concomitant cognitive and behavioural changes associated with development. We highlight some of the technical and conceptual challenges to be addressed by future developments in this rapidly moving field. Given the parallels previously discovered between neural systems across species and over a range of spatial scales, we also review some recent advances in developmental network studies at the cellular scale. We highlight the opportunities presented by such studies and how they may complement neuroimaging in advancing our understanding of brain development. Finally, we note that many brain and mind disorders are thought to be neurodevelopmental in origin and that charting the trajectory of brain network changes associated with healthy development also sets the stage for understanding abnormal network development. We therefore briefly review the clinical relevance of network metrics as potential diagnostic markers and some recent efforts in computational modelling of brain networks which might contribute to a more mechanistic understanding of neurodevelopmental disorders in future. © 2014 The Authors. Journal of Child Psychology and Psychiatry published by John Wiley & Sons Ltd on behalf of Association for Child and Adolescent Mental Health.
We examined fluctuations in band-limited power (BLP) of local field potential (LFP) signals recorded from multiple electrodes in visual cortex of the monkey during different behavioral states. We asked whether such signals demonstrated coherent fluctuations over time-scales of seconds and minutes, and would thus serve as good candidates for direct comparison with data obtained from functional magnetic resonance imaging (fMRI). We obtained the following results. (i) The BLP of the local field displayed fluctuations at many time-scales, with particularly large amplitude at very low frequencies (<0.1 Hz). (ii) These fluctuations exhibited high coherence between electrode pairs, particularly for BLP signals derived from the gamma (γ) frequency range. (iii) Coherence in the BLP, unlike that in the raw LFP, did not fall off sharply as a function of cortical distance. (iv) The structure and coherence of BLP changes were highly similar under distinctly different behavioral states. These results demonstrate the existence of widespread coherent activity fluctuations in the brain of the awake monkey over very long time-scales. We propose that such signals may make a significant contribution to the high variability observed in the time course of physiological signals, including those measured with functional imaging techniques. The results are discussed in the context of combined fMRI/electrophysiological recordings.
Pathological perturbations of the brain are rarely confined to a single locus; instead, they often spread via axonal pathways to influence other regions. Patterns of such disease propagation are constrained by the extraordinarily complex, yet highly organized, topology of the underlying neural architecture; the so-called connectome. Thus, network organization fundamentally influences brain disease, and a connectomic approach grounded in network science is integral to understanding neuropathology. Here, we consider how brain-network topology shapes neural responses to damage, highlighting key maladaptive processes (such as diaschisis, transneuronal degeneration and dedifferentiation), and the resources (including degeneracy and reserve) and processes (such as compensation) that enable adaptation. We then show how knowledge of network topology allows us not only to describe pathological processes but also to generate predictive models of the spread and functional consequences of brain disease.
Background Functional connectivity MRI is an emerging technique that can be used to investigate typical and atypical brain function in developing and aging populations. Despite some of the current confounds in the field of functional connectivity MRI, the translational potential of the technique available to investigators may eventually be used to improve diagnosis, early disease detection, and therapy monitoring.Method and ScopeBased on a comprehensive survey of the literature, this review offers an introduction of resting-state functional connectivity for new investigators to the field of resting-state functional connectivity. We discuss a brief history of the technique, various methods of analysis, the relationship of functional networks to behavior, as well as the translational potential of functional connectivity MRI to investigate neuropsychiatric disorders. We also address some considerations and limitations with data analysis and interpretation.Conclusions The information provided in this review should serve as a foundation for investigators new to the field of resting-state functional connectivity. The discussion provides a means to better understand functional connectivity and its application to typical and atypical brain function.
The vast majority of mental illnesses can be conceptualized as developmental disorders of neural interactions within the connectome, or developmental miswiring. The recent maturation of pediatric in vivo brain imaging is bringing the identification of clinically meaningful brain-based biomarkers of developmental disorders within reach. Even more auspicious is the ability to study the evolving connectome throughout life, beginning in utero, which promises to move the field from topological phenomenology to etiological nosology. Here, we scope advances in pediatric imaging of the brain connectome as the field faces the challenge of unraveling developmental miswiring. We highlight promises while also providing a pragmatic review of the many obstacles ahead that must be overcome to significantly impact public health.
MRI connectomics is an emerging approach to studying the brain as a network of interconnected brain regions. Understanding and mapping the development of the MRI connectome may offer new insights into the development of brain connectivity and plasticity, ultimately leading to improved understanding of normal development and to more effective diagnosis and treatment of developmental disorders. In this review, we describe the attempts made to date to map the whole-brain structural MRI connectome in the developing brain and pay a special attention to the challenges associated with the rapid changes that the brain is undergoing during maturation. The two main steps in constructing a structural brain network are i) choosing connectivity measures that will serve as the network "edges" and ii) finding an appropriate way to divide the brain into regions that will serve as the network "nodes". We will discuss how these two steps are usually performed in developmental studies and the rationale behind different strategies. Changes in local and global network properties that have been described during maturation in neonates and children will be reviewed, along with differences in network topology between typically and atypically developing subjects, for example, due to premature birth or hypoxic ischemic encephalopathy. Finally, future directions of connectomics will be discussed, addressing important steps necessary to advance the study of the structural MRI connectome in development.
The Human Connectome Project consortium led by Washington University, University of Minnesota, and Oxford University is undertaking a systematic effort to map macroscopic human brain circuits and their relationship to behavior in a large population of healthy adults. This overview article focuses on progress made during the first half of the 5-year project in refining the methods for data acquisition and analysis. Preliminary analyses based on a finalized set of acquisition and preprocessing protocols demonstrate the exceptionally high quality of the data from each modality. The first quarterly release of imaging and behavioral data via the ConnectomeDB database demonstrates the commitment to making HCP datasets freely accessible. Altogether, the progress to date provides grounds for optimism that the HCP datasets and associated methods and software will become increasingly valuable resources for characterizing human brain connectivity and function, their relationship to behavior, and their heritability and genetic underpinnings.