NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Contents lists available at ScienceDirect
NeuroImage
journal homepage: www.elsevier.com/locate/neuroimage
Imaging structural covariance in the development of intelligence
Budhachandra S. Khundrakpam a,n, John D. Lewis a, Andrew Reid b, Sherif Karama a,
Lu Zhao a, Francois Chouinard-Decorte a, Alan C. Evans a, Brain Development Cooperative
Group1
a
b
McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University, Montreal, Canada
Institute of Neuroscience and Medicine (INM-1), Research Centre Jülich, Jülich, Germany
art ic l e i nf o
a b s t r a c t
Article history:
Received 29 March 2016
Accepted 19 August 2016
Verbal and non-verbal intelligence in children is highly correlated, and thus, it has been difficult to
differentiate their neural substrates. Nevertheless, recent studies have shown that verbal and non-verbal
intelligence can be dissociated and focal cortical regions corresponding to each have been demonstrated.
However, the pattern of structural covariance corresponding to verbal and non-verbal intelligence remains unexplored. In this study, we used 586 longitudinal anatomical MRI scans of subjects aged 6–18
years, who had concurrent intelligence quotient (IQ) testing on the Wechsler Abbreviated Scale of Intelligence. Structural covariance networks (SCNs) were constructed using interregional correlations in
cortical thickness for low-IQ (Performance IQ ¼100 78, Verbal IQ¼ 100 77) and high-IQ (PIQ ¼121 7 8,
VIQ ¼120 79) groups. From low- to high-VIQ group, we observed constrained patterns of anatomical
coupling among cortical regions, complemented by observations of higher global efficiency and modularity, and lower local efficiency in high-VIQ group, suggesting a shift towards a more optimal topological organization. Analysis of nodal topological properties (regional efficiency and participation
coefficient) revealed greater involvement of left-hemispheric language related regions including inferior
frontal and superior temporal gyri for high-VIQ group. From low- to high-PIQ group, we did not observe
significant differences in anatomical coupling patterns, global and nodal topological properties. Our
findings indicate that people with higher verbal intelligence have structural brain differences from
people with lower verbal intelligence – not only in localized cortical regions, but also in the patterns of
anatomical coupling among widely distributed cortical regions, possibly resulting to a system-level reorganization that might lead to a more efficient organization in high-VIQ group.
& 2016 Published by Elsevier Inc.
Keywords:
Verbal and performance intelligence
Cortical thickness
Structural covariance network
Neuroimaging
Cognitive development
Introduction
One of the most thought-provoking questions we face is why
some people are more proficient in several cognitive skills than
others. Understanding the biological bases of these differences is
of utmost importance to basic and applied neuroscience. Initial
insights to the biological bases came from studies based on post
mortem data that revealed positive association between cerebral
volume and intelligence (Witelson et al., 2006). However, the advent of advanced MRI techniques which enabled scientists to investigate highly-localized (voxel-level) relationships of brain
measurements (e.g. gray matter density, cortical thickness) with
intelligence, provided a different perspective. Several such studies
at the voxel and regional levels demonstrated positive correlation
n
Corresponding author.
E-mail address: budha@bic.mni.mcgill.ca (B.S. Khundrakpam).
1
See Appendix for author list and affiliations of the Brain Development Cooperative Group.
of morphometry with intelligence in brain regions that are especially relevant to higher cognitive functions including frontal,
temporal, parietal, hippocampus and cerebellum (Andreasen et al.,
1993; MacLullich et al., 2002; Shaw et al., 2006; Narr et al., 2007;
Colom et al., 2009; Karama et al., 2011; Burgaleta et al., 2014).
Thus, came the proposition that increased volume in specific brain
regions may account for the association between intelligence and
global brain volume.
General intelligence is considered to be broadly dissociable into
fluid and crystalized intelligence (Cattell, 1943), and brain areas
corresponding to each have been shown in several studies (Choi
et al., 2008; Karama et al., 2011; Ramsden et al., 2011; Colom et al.,
2013; Burgaleta et al., 2014). Fluid intelligence, alternatively described as reasoning (non-verbal) ability, involves reasoning and
novel problem-solving ability (Cattell, 1943), and has been shown
to depend on working memory (Kyllonen and Christal, 1990; Kane
and Engle, 2002). Crystallized intelligence, on the other hand, refers to verbal ability; this includes the ability of using language in
analysing, remembering and understanding information, and is
http://dx.doi.org/10.1016/j.neuroimage.2016.08.041
1053-8119/& 2016 Published by Elsevier Inc.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
2
assumed to depend on acquired and accumulated knowledge including semantic memory (Naglieri and Bornstein, 2003). Apart
from the conceptual difference, verbal and non-verbal intelligence
have also been shown to be empirically separable. For example,
patients with lesions specifically in prefrontal cortex have lower
non-verbal intelligence while verbal intelligence is compromised
in patients with lesions specifically in anterior temporal regions
(Duncan et al., 1996; Waltz et al., 1999). The Wechsler Abbreviated
Scale of Intelligence (WASI) is used as a screener of verbal and
non-verbal abilities, and give verbal and performance IQ scores
(Wechsler, 1999). Using the VIQ and PIQ scores, several neuroimaging studies have also noted a dissociation of verbal and nonverbal intelligence: positive associations of GM density in temporal regions with verbal intelligence (Choi et al., 2008; Lee et al.,
2014) and in prefrontal regions with non-verbal intelligence (Gray
et al., 2003). It may be noted that VIQ and PIQ scores are partial
estimates and may not fully describe the verbal and non-verbal
abilities.
Although the above findings demonstrate dissociation of cortical regions for verbal and non-verbal intelligence, the possibility
that the relationships between different cortical regions vary in
distinct patterns for verbal and non-verbal intelligence remains
unexplored. Such a motivation arises in light of recent studies that
have revealed distinct patterns in anatomical coupling among
cortical regions associated with greater general intelligence (Lerch
et al., 2006) and vocabulary abilities (Lee et al., 2014). Lerch et al.
(2006) found stronger anatomical coupling between a seed region
at BA 44 (part of Broca’a area) with several frontal and parietal
regions for individuals with higher general intelligence; while Lee
et al. (2014) observed stronger anatomical coupling in multiple
regions involved in language in people who possess greater vocabulary skills. Although these two studies have shown modifications in anatomical coupling with greater general intelligence
and vocabulary, there has not been any study that explored the
dissociation of verbal and non-verbal intelligence in terms of differences in the patterns of anatomical coupling among cortical
regions. Given the distinct focal differences in cortical thickness
with greater verbal and non-verbal intelligence, we postulate that
there will be distinct variations in the patterns of anatomical
coupling with greater verbal and non-verbal intelligence.
A recently introduced methodology to examine anatomical
coupling among broadly distributed cortical regions instead of
focussing on each cortical region in statistical isolation from all
others, is the study of structural covariance networks (SCNs) (He
et al., 2007; Khundrakpam et al., 2013; for detail reviews, see
Alexander-Bloch et al., 2013; Evans, 2013). Several studies have
replicated patterns of SCNs in normal brains (Zielinski et al., 2010;
Raznahan et al., 2011; Khundrakpam et al., 2013), and alterations
in the SCN patterns have been shown in several diseases including
Alzheimer's disease, schizophrenia, multiple sclerosis, autism etc.
(Bassett and Bullmore, 2009; He et al., 2009; Sharda et al., 2014).
Additionally, accumulating evidence have also shown anatomical
(white matter connectivity) and functional (resting state fMRI
connectivity) correspondence with SCNs suggesting that SCN
patterns might capture some aspects of brain connectivity (Gong
et al., 2012; Kelly et al., 2012).
Given that SCNs provide a good framework for investigating
anatomical coupling among cortical regions, we hypothesize that
they will provide information about differences in the patterns of
anatomical coupling among cortical regions associated with
greater verbal and performance intelligence. Additionally, we aim
to explore differences in topological organization corresponding to
greater verbal and performance intelligence.
Materials and methods
Participants
Data for the study were taken from the NIH MRI Study of
Normal Brain Development (Evans and Brain Development Cooperative, 2006); a multi-site project undertaken to offer a normative database for normal brain and cognitive development. 586
MRI scans of subjects aged 6–18 years scanned up to 3 times at
2 year intervals that had concurrent intelligence quotient (IQ)
testing on the Wechsler Abbreviated Scale of Intelligence (WASI)
were used. Detailed demographics of the subjects are given in
Table 1.
Psychometric measures
Several batteries of behavioral measures were acquired from
the subjects on or within few days of brain imaging (for details,
see Evans and Brain Development Cooperative, 2006; Waber et al.,
2007). Cognitive measures used in the study were the Wechsler
Abbreviated Scale of Intelligence (WASI); (Wechsler, 1999) from
the NIH MRI Study of Normal Brain Development. The WASI consisted of vocabulary, similarities, matrix reasoning, and block design subtests. Verbal IQ (VIQ) and performance IQ (PIQ) measures
were computed by normalizing the scores on individual subtests
(vocabulary and similarities for VIQ, and matrix reasoning and
block design for PIQ) against age-specific norms. Thus, VIQ comprised those tests more related to verbal skills while PIQ involved
tests more independent of verbal skills.
MRI acquisition and processing
For each subject, a 3D T1-weighted (T1W) Spoiled Gradient
Recalled (SPGR) echo sequence with 1.5 T scanners was acquired,
with 1mm isotropic data obtained sagittally from the whole head.
For GE scanners, slice thickness of 1.5 mm was obtained due to
their limit of 124 slices. Additionally using a two-dimensional (2D)
Table 1
Demographics of the subjects used in the study. Means with standard deviation, and range given in parentheses. The last column shows the overlap/statistical dependency
between PIQ and VIQ scores. PIQ, performance IQ; VIQ, verbal IQ.
PIQ
VIQ
Group
Scans
Subjects (M/F)
Age
PIQ score
Subjects (M/F)
Age
VIQ score
t test (PIQ/VIQ)
Low-IQ
High-IQ
293
293
187 (88/99)
180 (86/94)
12.8 73.
12.9 73.8
1007 8 (72–110)
1217 8 (111–157)
189 (93/96)
185 (88/97)
13.0 7 3.7
12.7 7 3.7
100 77 (74–109)
1207 9 (110–156)
p ¼0.55
p ¼0.21
Total number of subjects, n ¼306 (scanned up to 3 times).
Total number of scans, N ¼ 586.
Males/females ¼ 141/165.
Age¼ 6–18 years.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
multi-slice (2 mm) dual echo fast spin echo (FSE) sequence, T2weighted (T2W) and proton density-weighted (PDW) images were
obtained. The total acquisition time was 25 min and was repeated when specified by the scanner-side quality control process.
For subjects that could not endured this procedure, a fallback
protocol with shorter 2d acquisitions of slice thickness 3 mm was
used.
Cortical thickness measurements
T1-weighted MRI scans were used to obtain cortical thickness
measures using the CIVET pipeline developed at the Montreal
Neurological Institute (http://www.bic.mni.mcgill.ca/ServicesSoft
ware/CIVET). The steps involved are given as follows. First, the
native MRI images were corrected for non-uniformity artifacts
using N3 algorithm (Sled et al., 1998) and registered into the stereotaxic space (Talairach and Tournoux, 1988) using a 9-parameter
linear transformation (Collins et al., 1994). The images were then
segmented into gray matter, white matter, cerebrospinal fluid and
background using an advanced neural net classifier (Zijdenbos
et al., 2002). Next, inner and outer gray matter surfaces were automatically extracted from each MR volume using the Constrained
Laplacian-based Automated Segmentation with Proximities
(CLASP) algorithm (MacDonald et al., 2000; Kim et al., 2005). Then,
cortical thickness was computed in native space using the linked
distance between the two surfaces at 81,924 vertices (163,840
polygons) throughout the cortex. Using manual measurements
(Kabani et al., 2001) and simulation approaches (Lerch and Evans,
2005; Lerch et al., 2005), the validation of the CLASP cortical
thickness algorithm, and has also been recently applied to Alzheimer's disease (Lerch et al., 2005) and cortical development
studies (Shaw et al., 2006).
Correlation strength maps
The data were divided into two groups with equal number of
scans in each (n¼ 293); the group with the lower PIQ scores was
defined as low-PIQ group and the other with higher PIQ scores as
high-PIQ group. Similarly, the group with the lower VIQ scores was
defined as low-VIQ group and the other with higher VIQ scores as
high-VIQ group. Details are given in Table 1. Since the data comprised of longitudinal MRI scans, a mixed effects general linear
model (GLM) was used to account for repeated subjects. Age and
gender were put as covariates in the mixed effects GLM for each
group, and the resulting residuals were used in all the analyses.
Next, correlation strength maps of cortical thickness showing
anatomical coupling among cortical regions at whole brain level
were computed for each group (Lerch et al., 2006; Raznahan et al.,
2011). As explained in earlier papers (Lerch et al., 2006; Raznahan
et al., 2011), correlation strength maps were calculated by computing the degree to which any part of the cortex correlated with
the rest of the cortex. This requires correlating cortical thickness at
each vertex with all the rest of vertices (n ¼81,923) and therefore
the process (81,924 81,924 number of correlations) is computationally expensive. Instead, a less computationally expensive approach which has been validated in previous studies (Lerch et al.,
2006; Raznahan et al., 2011; Lee et al., 2014), is to correlate cortical
thickness at each vertex with the mean cortical thickness (henceforth mCT which is calculated by averaging the cortical thickness
across all vertices within each subject). Thus, correlation strength
maps corresponding to low- and high-IQ groups were computed.
Network analysis
For network analysis, regions of interest (ROIs) were created
based on automatic anatomical labeling (AAL) scheme, which is
Table 2
List of brain parcels based on automatic anatomical labeling (AAL) atlas.
ROI-abbreviations ROI-labels
REC.L
OLF.L
ORBsup.L
ORBsupmed.L
ORBmid.L
ORBinf.L
SFGdor.L
MFG.L
IFGoperc.L
IFGtriang.L
SFGmed.L
SMA.L
PCL.L
PreCG.L
ROL.L
PoCG.L
SPG.L
IPL.L
SMG.L
ANG.L
PCUN.L
SOG.L
MOG.L
IOG.L
CAL.L
CUN.L
LING.L
FFG.L
HES.L
STG.L
MTG.L
ITG.L
TPOsup.L
TPOmid.L
PHG.L
ACG.L
DCG.L
PCG.L
INS.L
REC.R
OLF.R
ORBsup.R
ORBsupmed.R
ORBmid.R
ORBinf.R
SFGdor.R
MFG.R
IFGoperc.R
IFGtriang.R
SFGmed.R
SMA.R
PCL.R
PreCG.R
ROL.R
PoCG.R
SPG.R
IPL.R
SMG.R
ANG.R
PCUN.R
SOG.R
MOG.R
IOG.R
CAL.R
CUN.R
LING.R
FFG.R
HES.R
STG.R
MTG.R
ITG.R
TPOsup.R
TPOmid.R
PHG.R
ACG.R
Left Gyrus Rectus
Left Olfactory Cortex
Left Supeiror frontal gyrus, orbital part
Left Superior frontal gyrus, medial orbital
Left Middle frontal gyrus orbital part
Left Inferior frontal gyrus, orbital part
Left Superior frontal gyrus, dorsolateral
Left Middle frontal gyrus
Left Inferior frontal gyrus, opercular part
Left Inferior frontal gyrus, triangular part
Left Superior frontal gyrus, medial
Left Supplementary motor area
Left Paracentral lobule
Left Precentral gyrus
Left Rolandic operculum
Left Postcentral gyrus
Left Superior parietal gyrus
Left Inferior parietal, but supramarginal and angular gyri
Left Supramarginal gyrus
Left Angular gyrus
Left Precuneus
Left Superior occipital gyrus
Left Middle occipital gyrus
Left Inferior occipital gyrus
Left Calcarine fissure and surrounding cortex
Left Cuneus
Left Lingual gyrus
Left Fusiform gyrus
Left Heschl gyrus
Left Superior temporal gyrus
Left Middle temporal gyrus
Left Inferior temporal gyrus
Left Temporal pole: superior temporal gyrus
Left Temporal pole: middle temporal gyrus
Left Parahippocampal gyrus
Left Anterior cingulate and paracingulate gyri
Left Median cingulate and paracingulate gyri
Left Posterior cingulate gyrus
Left Insula
Right Gyrus Rectus
Right Olfactory Cortex
Right Superior frontal gyrus, orbital part
Right Superior frontal gyrus, medial orbital
Right Middle frontal gyrus orbital part
Right Inferior frontal gyrus, orbital part
Right Superior frontal gyrus, dorsolateral
Right Middle frontal gyrus
Right Inferior frontal gyrus, opercular part
Right Inferior frontal gyrus, triangular part
Right Superior frontal gyrus, medial
Right Supplementary motor area
Right Paracentral lobule
Right Precentral gyrus
Right Rolandic operculum
Right Postcentral gyrus
Right Superior parietal gyrus
Right Inferior parietal, but supramarginal and angular gyri
Right Supramarginal gyrus
Right Angular gyrus
Right Precuneus
Right Superior occipital gyrus
Right Middle occipital gyrus
Right Inferior occipital gyrus
Right Calcarine fissure and surrounding cortex
Right Cuneus
Right Lingual gyrus
Right Fusiform gyrus
Right Heschl gyrus
Right Superior temporal gyrus
Right Middle temporal gyrus
Right Inferior temporal gyrus
Right Temporal pole: superior temporal gyrus
Right Temporal pole: middle temporal gyrus
Right Parahippocampal gyrus
Right Anterior cingulate and paracingulate gyri
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
3
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
4
Table 2 (continued )
ROI-abbreviations ROI-labels
DCG.R
PCG.R
INS.R
Right Median cingulate and paracingulate gyri
Right Posterior cingulate gyrus
Right Insula
widely used for brain parcellation (Tzourio-Mazoyer et al., 2002).
Since our analysis is based on a cortical surface model, the AAL
atlas resulted into 78 cortical regions (Table 2). For every brain
region, cortical thickness was then computed as the average
thickness of all vertices defined as belonging to that region
(Khundrakpam et al., 2013).
In the next step, correlation matrices were constructed on the
basis of correlations of cortical thickness for a group of subjects
corresponding to low- and high-IQ groups (He et al., 2007;
Khundrakpam et al., 2013). For every region, a linear regression
was used to remove the effects of age, gender and mean overall
cortical thickness. The resulting residuals were used to compute
the statistical similarity in cortical thickness between any two
brain regions across a group of subjects, resulting to a 78 78
correlation matrix. We thus obtained the following correlation
matrices –
PIQ
PIQ
i) CMLow
and CMHigh
for low- and high-PIQ groups respectively.
VIQ
VIQ
and CMHigh
for low- and high-VIQ groups respectively.
ii) CMLow
For graph-theoretic analysis, the correlation matrices were first
binarized based on sparsity thresholding. We defined sparsity as
the total number of edges in a graph divided by the maximum
possible number of edges. Fixing a sparsity threshold (for instance,
x% threshold denotes x% of the topmost connections) assured the
same number of edges for graphs of different groups (Achard and
Bullmore, 2007). In order to determine proper assessment of small
world network properties and nominal spurious number of edges
in each network, we used a range of sparsity threshold 10 rS r30,
as has been done in earlier studies (Bassett et al., 2008; Khundrakpam et al., 2013). Since the exact sparsity of brain connections
is not clearly known, and the fact that group difference in network
topology may occur at different sparsities (He et al., 2008), a
metric summarizing information across sparsities is used. In the
present study, one such summary metric which is the normalized
integral of the network parameter across the sparsity range (He
et al., 2009a; Khundrakpam et al., 2013), was used for the topological properties.
The binarized graphs were used to calculate i) global network
parameters; namely, local and global efficiency, and modularity
and ii) local network parameters namely, regional efficiency and
participation coefficient which are described below.
Local and global efficiency are distance-weighted measures
(following Latora and Marchiori, 2001), and are defined as
Eglobal =
Elocal =
1
N (N−1)
1
N
∑
i≠j∈G
1
dij
∑ Eglobal (Gj )
j∈G
where G is the network graph, i and j are ROIs, dij is a measure of
distance between i and j (obtained from Euclidean distance).
Modularity is a measure of the degree to which a graph is organized into distinct subgraphs (Newman, 2006). The modularity
algorithm attempts to partition the graph into disjoint subgraphs,
in such a way as to minimize the degree of intermodule connectivity, and maximize intramodule connectivity. This ratio is
captured in the parameter Q, and for the study, we used the
maximal value Qmax.
In addition to the global modularity index, we also analyzed
nodal modular structure, namely participation coefficient (PC). PC
indicates the density of connections of a node i to nodes in other
modules (Guimera et al., 2005). Thus, nodes with high PC are
normally ‘connector hubs’ that integrate information between
disparate brain sub-systems (Chen et al., 2008; He et al., 2009b).
We define regional efficiency, Eregional (i), of a node i as the inverse of the harmonic mean of the shortest path length between
the node i and rest of the nodes in the graph (Achard et al., 2006).
A node with high regional efficiency reflects large number of
connections passing through the node, and thus is considered a
‘hub’ node (Achard et al., 2006; Gong et al., 2009).
Statistical analysis
Group difference in cortical thickness between low- and highIQ groups was analyzed using a vertex-wise mixed effects linear
model in which age, gender were taken into account. T-statistics
with random field theory (RFT) for correction of multiple comparisons, were used to quantify the group difference in cortical
thickness using SurfStat (http://www.math.mcgill.ca/keith/surf
stat/). We also computed group difference in lobar cortical thickness for the IQ groups. Residuals of cortical thickness values (after
removing the effects of age and gender) for all vertices belonging
to a lobe for low-IQ were compared to that of high-IQ group using
Student's t-test. Adjustment for multiple comparisons was done at
FDR of q ¼0.05.
Group difference in the correlation strength maps between
low- and high-IQ groups were computed in a procedure similar to
earlier studies (He et al., 2008). Correlating cortical thickness at
each vertex with the mean cortical thickness (mCT) results in a
correlation coefficient for each vertex. The correlation coefficients
were converted into z values based on Fisher's r to z transform.
The transformed z values were used to compute a Z statistic which
in turn allowed statistical comparison of the group difference in
the correlations (Cohen and Cohen, 1983). Correction for multiple
comparisons was done using a false discovery rate (FDR) at
q¼ 0.05 (Genovese et al., 2002).
Statistical comparison of the network properties between the IQ
groups was performed using a nonparametric permutation test procedure (Bullmore et al., 1999; He et al., 2008). For the low- and high-IQ
groups, the network properties were first calculated separately for the
range of sparsity thresholds and the summary metrics (normalized
integrals) obtained. Next, in order to test the null hypothesis that
group differences might occur due to chance, each subject's set of
cortical thickness values were randomly reallocated to one or other of
the two groups. Then, correlation matrix was computed for each
randomized group. Following similar procedure as in network construction, the randomized correlation matrix was used to compute
network properties across the range of sparsity thresholds. Next, the
differences in the normalized integrals of the network properties for
the randomized groups were obtained. We repeated the randomization procedure 1000 times and used the 95 percentile points for each
distribution as the critical values for a one-tailed test of the null hypothesis with a probability of type I error of 0.05. In the study, the
above mentioned procedure was used to compare group differences in
all global (global and local efficiency, modularity) and nodal (regional
efficiency and participation coefficient) topological properties. For the
nodal properties, a region-by-region comparison was performed using
the nonparametric permutation. Adjustment for multiple comparisons
was done using a false discovery rate (FDR) at q¼0.05.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
5
Fig. 1. Group difference in cortical thickness between Low- and High-IQ groups. Significant group differences (High–Low) in cortical thickness (p o 0.05, random field theory,
RFT-corrected) are depicted on the surface of a brain template. A and B are for PIQ and VIQ groups respectively.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
6
Results
Group difference in cortical thickness between low- and high-IQ
groups
Comparison of group difference in cortical thickness revealed
greater thickness in several brain regions for both the high-PIQ
and high-VIQ groups (Fig. 1). In particular, significantly greater
cortical thickness (p o0.05, RFT-corrected) was observed for the
high-PIQ group in several regions, including left inferior temporal
gyrus, right middle temporal gyrus, bilateral cuneus, right superior
occipital and left inferior occipital gyri, bilateral anterior cingulate
gyri, bilateral insula, right inferior frontal and precentral gyri
(Fig. 1A and Table 3A). We also observed significantly greater
cortical thickness (p o0.05, RFT-corrected) for the high-VIQ group
in bilateral middle temporal gyri, left inferior temporal gyrus, bilateral precuneus, left cuneus, left lingual and parahippocampal
gyri, left inferior frontal gyrus, bilateral insula and right superior
occipital and superior parietal gyri (Fig. 1B and Table 3B).
Significant differences in lobar cortical thickness were observed
between low- and high-VIQ groups (Fig. 2). Cortical thickness was
significantly greater for high-VIQ (p o0.05, FDR-corrected) in occipital, temporal, and limbic cortex (second row of Fig. 2). There
were no significant differences in any lobar cortical thickness between PIQ groups (first row of Fig. 2).
and high-IQ groups was determined (Fig. 3). No significant difference in
correlation strength maps between the low- and high-PIQ groups was
observed (Fig. 3A). However, we observed significantly greater correlation strengths in bilateral temporal, inferior frontal, and lateral parietal regions for the low-VIQ (compared to the high-VIQ group)
(Fig. 3B).
Group difference in global topological properties between low- and
high-IQ groups
Comparative analysis (nonparametric permutation tests) of
global topological properties including global and local efficiency,
and modularity were performed between the low- and high-IQ
groups (Fig. 4). We observed significantly higher global efficiency
and modularity (po 0.05), and significantly lower local efficiency
(p o0.05) for the high- compared to low-VIQ groups (Fig. 4). There
were however no significant differences in the global topological
parameters between the low- and high-PIQ groups (Fig. 4).
Group difference in correlation strength maps between low- and
high-IQ groups
Correlation strength maps were computed for all IQ groups, and
then group difference in the correlation strength maps between lowTable 3
Cortical regions with significant group difference (p o 0.05, RFT-corrected at cluster
level) in cortical thickness for A. (High–Low) PIQ and B. (High–Low) VIQ.
A (High–Low) PIQ
Cluster ID
No. of vertices
P-value
Brain label
1
2
3
4
5
6
7
8
9
10
11
12
13
14
908
474
528
703
477
359
434
299
313
244
167
101
139
178
1.00E 07
1.07E 07
1.23E 06
2.35E 06
4.35E 06
4.33E 05
0.000239
0.00041
0.001727
0.010569
0.012696
0.015828
0.035413
0.03928
INS.R
IFGtriang.R
CUN.R
ACG.L
INS.L
ITG.L
CUN.L
ACG.R
SOG.R
IOG.L
ITG.L
PreCG.R
MTG.R
STG.R
Cluster ID
No. of vertices
P-value
Brain label
1
2
3
4
5
6
7
8
9
10
11
12
13
775
1608
546
611
184
647
443
310
249
273
220
157
91
1.00E 07
1.00E 07
1.03E 07
1.82E 07
4.28E 06
5.66E 06
0.0002007
0.0034167
0.0078855
0.0080145
0.0114672
0.0122702
0.0153094
PCUN.R
CUN.L
MTG.L
PHG.L
INS.R
LING.L
PCUN.L
IFGtriang.L
ITG.L
INS.L
SOG.R
SPG.R
MTG.R
B (High–Low) VIQ
Fig. 2. Comparison of lobar cortical thickness (residuals) for Low- and High-IQ
groups. Residual cortical thickness (after controlling for age, gender) were compared between low- and high-IQ groups using Student t-tests. Fig. 3A and B are for
PIQ and VIQ groups respectively. Note the significant increase in cortical thickness
in high- compared to low-VIQ group for occipital, temporal and limbic lobes.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
7
Fig. 3. Comparison of correlation strength maps for Low- and High-IQ groups. Correlation strength maps representing overall connectivity are shown in the first column,
upper for low-IQ and lower for high-IQ groups. The second column shows the statistical comparison of correlation strength maps at p o 0.05 (FDR-corrected for multiple
comparisons) for low minus high IQ groups. A and B are for PIQ and VIQ groups respectively. Note the color bar in the right column of B shows Z statistics for the group
difference (low minus high VIQ) (see Method).
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
8
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Fig. 4. Comparison of global topological properties for Low- and High-IQ groups. Global and local efficiency, and modularity were compared between low- and high-IQ
groups using two approaches, parametric (A) and non-parametric (B) (see Methods for details). In A and B, the first row shows comparison for low- and high-PIQ groups,
while the second row shows comparison for low- and high-VIQ groups. Note that both the approaches yielded the same result: significantly greater global efficiency,
modularity and significantly lower local efficiency for high-VIQ compared to low-VIQ group.
low-VIQ (Fig. 5). There were no cortical regions with significant
difference in PC metric between low- and high-PIQ groups.
Region-by-region (regions based on AAL parcellation) statistical
comparison of regional efficiency using the nonparametric permutation test revealed 3 cortical regions with significant difference (p o0.05, FDR-corrected) for high-VIQ compared to low-VIQ
(Fig. 6B). Note that, for better visualization of all comparisons for
78 cortical regions, z-scores for the difference in regional efficiency
have been shown on x-axis (with * for regions with significant
difference). There were no cortical regions with significant difference in regional efficiency between low- and high-PIQ groups
(Fig. 6A).
Discussion
Fig. 5. Comparison of connector hubs for Low- and High-IQ groups. Based on
participation coefficient, connector hubs were identified for the IQ groups and
compared (see Methods for details). Five brain regions were identified as connector
hubs with significant difference in participation coefficient between low- and highVIQ groups. Similar analysis for PIQ groups yielded no significant difference in
participation coefficient. Note, IFGoperc.L¼ left inferior frontal gyrus, opercular
part; SMG.L ¼ left supramarginal gyrys; STG.L¼ left superior temporal gyrus; MTG.
L¼ left middle temporal gyrus and MTG.R ¼right middle temporal gyrus.
Group difference in nodal topological properties between low- and
high-IQ groups
Analysis of group difference in participation coefficient (PC)
revealed significantly greater (p o0.05, FDR-corrected) PC metric
predominantly in left-hemisphere including inferior frontal gyrus,
supra marginal gyrus, superior and temporal gyri and right
hemispheric middle temporal gyrus for high-VIQ compared to
In the present study, from low- to high-VIQ group, we found i)
different anatomical coupling among widely distributed cortical
regions; ii) a difference in global topological properties (higher
global efficiency and modularity, and lower local efficiency, suggesting a shift towards a more optimal topological organization);
and iii) a difference in regional efficiency for cortical regions predominantly in the left hemisphere. There were, however, no significant differences in SCNs and global topological properties between the low- and high-PIQ groups. Taken together, our results
show that people with higher verbal intelligence have structural
brain differences from people with lower verbal intelligence – not
only in localized cortical regions, but also in the patterns of anatomical coupling among widely distributed cortical regions. Similar conclusions could not be made for people with higher performance intelligence.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
9
Fig. 6. Comparison of regional efficiency for Low- and High-IQ groups. Group difference (High–Low) in regional efficiency are shown for all the 78 brain parcels (based on
automatic anatomical labeling (AAL) atlas, see Table 2 for the brain labels). A and B are for PIQ and VIQ groups respectively. In B, 3 brain regions were identified as regions
with significantly different regional efficiency (p o 0.05 denoted by *) between low- and high-VIQ groups (see Methods for details on the statistical comparison). Note, no
brain regions showed significant difference in regional efficiency for low- and high-PIQ groups (A).
Association of cortical thickness and intelligence
Early explorations of the biological substrates of intelligence
focussed on studies using post mortem data and demonstrated
positive association of cerebral volumes with intelligence (Witelson et al., 2006). However, the advent of automated tissue
classification with advanced MRI allowed scientists to investigate
highly-localized (voxel-level) relationships of intelligence with
brain measurements (Shaw et al., 2006; Narr et al., 2007; Karama
et al., 2011; Karama et al., 2014).
Our finding of greater cortical thickness in several cortical regions for both high PIQ and VIQ groups are in accordance with
earlier studies (Karama et al., 2011). These morphometric differences were observed in inferior frontal, temporal, and cingulate
cortex (Fig. 1A). Karama et al. (2011) found significant positive
correlations between block design (subtest of PIQ) and cortical
thickness in left frontal (BA 45), right frontal (BAs 8, 9, 10, 11, 46),
left parietal (BA 19), left temporal (BAs 20, 22, 36, 37, 38), right
temporal (20, 21, 36, 38) and left occipital (BA 19).
For verbal intelligence, we found significant group differences
in left frontal, left occipital, and temporal regions (Fig. 1B). Karama
et al. (2011) found significant positive correlations between vocabulary and peak coordinates at left inferior frontal gyrus (BA 45),
left precuneus (BA 7) and left occipito-temporal gyrus (BA 36). A
group study has also shown major increase in cortical thickness in
bilateral temporal cortex for both high- compared to low-VIQ
group (Choi et al., 2008).
Though several studies have shown positive correlation of cortical
thickness (or GM density) in specific brain regions with intelligence,
the mechanisms behind them are not clear. Greater cortical thickness
in certain brain regions for an enhanced intellectual ability might
suggest increased recruitment of those brain regions in that particular
intellectual ability. For example, high VIQ has been shown to be associated with greater cortical thickness in lateral temporal (Choi et al.,
2008) and posterior occipital cortex, regions which are involved in
visual and linguistic abilities. In particular, significant positive correlations of verbal intelligence ability (specifically, vocabulary size) with
GM density have been shown in bilateral posterior supramarginal
gyrus (SMG) (Lee et al., 2007), an anatomical site linked to second
language skill in adult bilinguals (Mechelli et al., 2004). Another study
also observed positive correlation of the acquisition of vocabulary
knowledge with GM density in left posterior SMG in teenagers. Thus
GM density in the posterior SMG appears to be sensitive to the
number of words learned, regardless of language. Similarly, greater
cortical thickness with high PIQ has been observed in lateral prefrontal
cortex, a region implicated in cognitive abilities related to tasks such as
planning, problem solving, reasoning (Duncan and Owen, 2000;
Duncan et al., 2000; Gray et al., 2003).
The cellular events leading to greater cortical thickness with
better intellectual abilities are not clear. Cortical thickness is believed to reflect the number of neurons per column along with
glial support and dendritic arborization (Rakic, 1988; Chklovskii
et al., 2004; la Fougere et al., 2011). The positive association of
cortical thickness and intellectual abilities are likely the outcome
of modifications in the amount of glial and capillary support as
well as in dendritic arborization (Chklovskii et al., 2004; Sur and
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
10
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Rubenstein, 2005), and due to variations in the number of cortical
neurons as there is no evidence for neuronal proliferation in the
human cortex during most of post-natal development (Zatorre
et al., 2012). This is consistent with findings that high IQ levels are
linked to greater number of dendrites (Jacobs and Scheibel, 1993)
while low IQ levels are associated with reduction in dendritic
branching (Huttenlocher, 1991). Increased dendritic arborization
has also been shown in studies involving animals with enriched
environmental experience (Diamond, 2001; Lerch et al., 2011).
Additionally, animal studies have shown experience- and learningrelated increases in the number of synapses and glial cells (Black
et al., 1990; Isaacs et al., 1992; Anderson et al., 1994; Anderson
et al., 1996; Kleim et al., 1996). However, there is the on-going
debate over the relative contribution of experience-dependent
anatomical plasticity versus pre-existing and pre-disposing anatomy to the differences for low and high IQ groups. For example,
positive correlation between vocabulary and GM density in posterior SMG in both first and second languages suggests that the
relative GM density increase reflects increased dendritic arborization or local synaptic density that results from experience dependent plasticity (Mechelli et al., 2004; Lee et al., 2007). At the
same time, there is the possibility of an alternate hypothesis that
people with greater skills for vocabulary acquisition may have a
genetic disposition in the structure of posterior SMG that might be
beneficial for vocabulary learning.
Anatomical coupling of cortical regions and intelligence
Our findings show significant difference in the correlation
strength maps between the low- and high-VIQ groups, but not for
PIQ groups (Fig. 2). Significant difference in correlation strengths
were observed at bilateral temporal and left lateral inferior frontal
regions (Fig. 2). The possible reasons behind these findings might
be explained as below.
A strong correlation coefficient reflects highly similar pattern of
cortical thickness between regions X and Y while a weak correlation coefficient demonstrates a dissimilar pattern. The results of
our study point to such dissimilarity for the high-VIQ group. In
both low-PIQ and low-VIQ groups as well as high-PIQ group, the
intra-lobar cortical thickness are not significantly different (Fig. 2).
Due to this, the correlation strength of any cortical region (which
reflects the summation of all the correlation coefficients with the
rest of the cortex) would be high (Fig. 3). However, in case of highVIQ group, there is significant increase in cortical thickness in
occipital, temporal and limbic cortex (Fig. 2B), thereby resulting in
a dissimilar pattern of cortical thickness and a corresponding decrease in the correlation strengths.
Our analysis of correlation strength maps was undertaken to
support and extend the findings of the low- versus high-IQ cortical
thickness contrast in discrete brain regions to differences in patterns of anatomical coupling among several brain regions with
varying levels of intelligence. Although the results might seem
contradictory to the notion of increased overall correlation
strength for high-IQ (Lerch et al., 2006), our results might however
be interpreted with the following hypothesis. In Fig. 3B, differences between low- and high-VIQ groups were observed in several
cortical regions. Thus, there was a higher number of regions in the
low-VIQ brains that were strongly correlated with each other. As
seen from Fig. 3B, there is a significant reduction of the correlation
strength in several regions for the high-VIQ group, demonstrating
a departure from the pattern seen for low-VIQ which might be
speculated as a selective process of specific regions and networks
that in turn reflects the plastic effects of verbal intelligence-related
activities. Taken together, the results indicate specialization of
cortical regions and possibly networks with higher verbal intelligence, similar to a phenomenon observed for musicians
(Bermudez et al., 2009). The findings from graph-theoretical
analyses discussed in following sections lend support to such a
hypothesis.
Differences in global topological properties
The analysis of global topological parameters revealed distinct
differences in topological organization between low- and high-VIQ
groups, but not between the PIQ groups. Significantly higher global efficiency and modularity as well as lower local efficiency for
high- compared to low-VIQ group suggests a shift of topological
organization towards a more optimal topological configuration.
Such a difference in topological organization was however, not
observed for high- compared to low-PIQ groups.
Graph theory has recently been used in several studies to understand normal and diseased brains (Bassett and Bullmore, 2009;
Bullmore and Sporns, 2009; He et al., 2009; He and Evans, 2010;
Evans, 2013; Sporns, 2013). One of the major advantages of using
graph-theoretical analysis in neuroscience is the ability to investigate whole-brain level exploration of brain networks that has
enabled neuroscientists to view diseased brains as ‘altered connectivity’ as opposed to traditional viewpoint of local disruptions
(He et al., 2009; Fornito and Bullmore, 2012; Hong et al., 2014; van
den Heuvel and Fornito, 2014; Fornito et al., 2015). Global, local
efficiency and modularity are examples of such graph parameters
which are used to explore global topological properties (Bullmore
and Sporns, 2009; Rubinov and Sporns, 2010; Sporns, 2013) and
have been used to understand global topological organization in
normal (Evans, 2013; Khundrakpam et al., 2013), and diseased
brains (Bassett and Bullmore, 2009; He et al., 2009). Of particular
interest to this study, these graph parameters have also been used
to understand IQ development using DTI (Li et al., 2009), fMRI (van
den Heuvel et al., 2009) and EEG (Langer et al., 2012). Irrespective
of the imaging modality, the studies found positive correlations of
FSIQ with global efficiency suggesting a more proficient brain organization with increased level of intelligence. Our results are
consistent with these earlier studies, and extend them by investigating IQ groups using structural covariance networks. We go
a step further and tease apart the topological differences between
different levels of verbal and performance intelligence. Consistent
with the findings of previous studies, we also observed significantly higher global efficiency and modularity as well as lower
local efficiency for the high- compared to low-VIQ group which
reflects a shift in topological organization towards a more optimal
configuration. However, we did not observe any significant changes in the topological parameters for low- to high-PIQ.
The findings are not surprising given the differences observed
in correlation strength maps between low- and high-VIQ groups,
but not between the PIQ groups. The graph-theoretical parameters
of global, local efficiency and modularity were computed from
correlation matrices of the IQ groups (see Methods). The correlation matrix is a low-dimensional construct (78 78 matrix based
on AAL template) which is a reduction from the high-dimensional
correlation strength maps (in 81,924 cortical regions at vertex level). Therefore, as seen in the case of correlation strength maps, it
is not unexpected that the correlation matrix belonging to highVIQ group is markedly different from that of the low-VIQ group. A
similar explanation holds for the observation of no significant
difference between the low- and high-PIQ groups. It may be noted
that the use of the low-dimensional construct (by averaging cortical thickness of vertices belonging to each ROI) is a potential
limitation of the study as focal changes in cortical thickness (at
vertex-level) have been shown to be associated with cognitive
changes (Shaw et al., 2006).
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
11
Differences in nodal topological properties
Conclusions
The findings of group difference in participation coefficient (PC)
and regional efficiency revealed significantly greater regional efficiency and PC specifically in left hemisphere language regions for
high-VIQ (compared to low-VIQ). The findings indicate left hemisphere dominance in brain's processing abilities with higher verbal intelligence. This observation is consistent with several structural and functional studies that have shown stronger correlations
of higher verbal skills with structure (cortical thickness/GM density) and function (activations) of left hemispheric language related regions facilitated by enhanced white matter connectivity
among the regions (Schmithorst et al., 2005). A strong bias toward
left-hemisphere representation of language skills have been
documented by lesion studies (Broca, 1861; Dronkers et al., 2004;
Dronkers et al., 2007) and neuroimaging studies (Bishop, 2013;
Gotts et al., 2013; Wang et al., 2014). Reading scores have been
found to be significantly correlated with white matter diffusion
anisotropy in the temporo-parietal region of the left hemisphere,
possibly contributing to reading ability by strengthening the
communication between cortical areas involved in visual, auditory,
and language processing (Klingberg et al., 2000). Evidence for this
has also come from studies on individuals with reading abnormality or dyslexia that have shown dysfunction of left hemisphere
reading networks (van der Mark et al., 2011; Richlan, 2012).
The left hemisphere representation of language concords well
with proposals that more spatially restricted, focal cortical representations that permit rapid cortical interactions with shorter
conduction delays, could be beneficial for systems that require
temporally rapid, sequential interactions (Semmes, 1968; Lackner
and Teuber, 1973; Poeppel, 2003) such as speech production and
comprehension (Hickok and Poeppel, 2007). In consistent with
this proposal, Gotts et al. (2013) observed that cortical regions in
the left hemisphere involved in language displayed a preference to
interact more exclusively with itself, and the magnitude of leftlateralization measured for individual subjects predicted the verbal ability levels.
Functional imaging studies have shown that individuals with
higher skill levels (e.g. good readers) tend to have more efficient
neural processes (generate less brain activity; than individuals
with lower skill levels (e.g. poor readers) (Maxwell et al., 1974;
Haier et al., 1988; Boivin et al., 1992; Reichle et al., 2000; Neubauer
and Fink, 2009). The interplay between brain structure and function being the root of the neural basis of cognition, learning and
plasticity; one can speculate whether the altered patterns of brain
activation leading to more efficient neural processing in higher
skill levels entail changes in brain structure (e.g. more synaptic
connections, more dendrites) or vice versa. Our results point to
such a hypothesis: larger cortical thickness in left frontal, left occipital and temporal regions for high-VIQ group might reflect
greater dendritic arborization in the cortical regions involved in
verbal skills. Such localized modifications of brain structure in
specific cortical regions for high VIQ group in turn, may be manifested as alterations of anatomical coupling among those regions,
indicating a system-level reorganization that might lead to a more
efficient organization in high-VIQ group.
Dissociation of verbal and non-verbal intelligence
Our findings of differences in anatomical coupling between the
verbal IQ groups, but not between the performance IQ groups
indicate dissociation of verbal and non-verbal intelligence, in
agreement with earlier studies (Choi et al., 2008; Colom et al.,
2009; Karama et al., 2011; Ramsden et al., 2011; Burgaleta et al.,
2014). Additionally, in partial agreement with our findings, there is
evidence indicating that greater verbal intelligence is more associated with changes in brain structure specifically in the left
hemisphere; while greater performance intelligence may be more
associated with changes in brain function (Choi et al., 2008). This
proposition is not completely unfounded and may largely be understood from the perspective of viewing cognition as a network
function. According to this viewpoint, performance intelligence
corresponds to the ability to dynamically configure information
processing network of cortical centers (predominantly prefrontal
cortex) to effectively deal with a novel cognitive challenge of some
complexity. Thus, performance intelligence may reflect the system's ability to dynamically bring together capabilities that are
distributed across different parts of the brain (Kaufman and
Kaufman, 1983; Naglieri and Das, 1997). On the contrary, verbal
intelligence is assumed to reflect achievement and not ability
per se, as performance on many of the verbal IQ tests, such as
vocabulary, depends on acquired knowledge (Naglieri and Bornstein, 2003) with predominant involvement of temporal regions.
Conflict of interest
The authors declare no conflict of interest.
Disclaimer
This manuscript reflects the views of the authors and may not
reflect the opinions or views of all Study Investigators or the NIH.
Acknowledgments
Funding: This research has been supported by The Azrieli
Neurodevelopmental Research Program in partnership with Brain
Canada Multi-Investigator Research Initiative (MIRI) (grant number PT-62570). BSK was supported by a Post-Doctoral Fellowship
from FRSQ and Jeanne-Timmins Costello MNI Fellowship.
This project has been funded in whole or in part with Federal
funds from the National Institute of Child Health and Human
Development, the National Institute on Drug Abuse, the National
Institute of Mental Health, and the National Institute of Neurological Disorders and Stroke (Contract #s N01-HD02-3343, N01MH9-0002, and N01-NS-9-2314, -2315, -2316, -2317, -2319 and
-2320). Special thanks to the NIH contracting officers for their
support.
We also acknowledge the important contribution and remarkable spirit of John Haselgrove, Ph.D. (deceased).
Appendix A. Brain Development Cooperative Group
Key personnel from the six pediatric study centers are as follows: Children's Hospital Medical Center of Cincinnati, Principal
Investigator William S. Ball, M.D., Investigators Anna Weber Byars,
Ph.D., Mark Schapiro, M.D., Wendy Bommer, R.N., April Carr, B.S.,
April German, B.A., Scott Dunn, R.T.; Children's Hospital Boston,
Principal Investigator Michael J. Rivkin, M.D., Investigators Deborah Waber, Ph.D., Robert Mulkern, Ph.D., Sridhar Vajapeyam, Ph.
D., Abigail Chiverton, B.A., Peter Davis, B.S., Julie Koo, B.S., Jacki
Marmor, M.A., Christine Mrakotsky, Ph.D., M.A., Richard Robertson,
M.D., Gloria McAnulty, Ph.D; University of Texas Health Science
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
12
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Center at Houston, Principal Investigators Michael E. Brandt, Ph.
D., Jack M. Fletcher, Ph.D., Larry A. Kramer, M.D., Investigators
Grace Yang, M.Ed., Cara McCormack, B.S., Kathleen M. Hebert, M.A.,
Hilda Volero, M.D.; Washington University in St. Louis, Principal
Investigators Kelly Botteron, M.D., Robert C. McKinstry, M.D., Ph.D.,
Investigators William Warren, Tomoyuki Nishino, M.S., C. Robert
Almli, Ph.D., Richard Todd, Ph.D., M.D., John Constantino, M.D.;
University of California Los Angeles, Principal Investigator James
T. McCracken, M.D., Investigators Jennifer Levitt, M.D., Jeffrey Alger,
Ph.D., Joseph O’Neil, Ph.D., Arthur Toga, Ph.D., Robert Asarnow, Ph.
D., David Fadale, B.A., Laura Heinichen, B.A., Cedric Ireland B.A.;
Children's Hospital of Philadelphia, Principal Investigators DahJyuu Wang, Ph.D. and Edward Moss, Ph.D., Investigators Robert A.
Zimmerman, M.D., and Research Staff Brooke Bintliff, B.S., Ruth
Bradford, Janice Newman, M.B.A. The Principal Investigator of the
data coordinating center at McGill University is Alan C. Evans, Ph.
D., Investigators Rozalia Arnaoutelis, B.S., G. Bruce Pike, Ph.D., D.
Louis Collins, Ph.D., Gabriel Leonard, Ph.D., Tomas Paus, M.D., Alex
Zijdenbos, Ph.D., and Research Staff Samir Das, B.S., Vladimir Fonov, Ph.D., Luke Fu, B.S., Jonathan Harlap, Ilana Leppert, B.E., Denise Milovan, M.A., Dario Vins, B.C., and at Georgetown University, Thomas Zeffiro, M.D., Ph.D. and John Van Meter, Ph.D. Ph.
D. Investigators at the Neurostatistics Laboratory, Harvard University/McLean Hospital, Nicholas Lange, Sc.D., and Michael P.
Froimowitz, M.S., work with data coordinating center staff and all
other team members on biostatistical study design and data analyses. The Principal Investigator of the Clinical Coordinating Center
at Washington University is Kelly Botteron, M.D., Investigators C.
Robert Almli Ph.D., Cheryl Rainey, B.S., Stan Henderson M.S., Tomoyuki Nishino, M.S., William Warren, Jennifer L. Edwards M.SW.,
Diane Dubois R.N., Karla Smith, Tish Singer and Aaron A. Wilber,
M.S. The Principal Investigator of the Diffusion Tensor Processing
Center at the National Institutes of Health is Carlo Pierpaoli, MD,
Ph.D., Investigators Peter J. Basser, Ph.D., Lin-Ching Chang, Sc.D.,
Chen Guan Koay, Ph.D. and Lindsay Walker, M.S. The Principal
Collaborators at the National Institutes of Health are Lisa Freund,
Ph.D. (NICHD), Judith Rumsey, Ph.D. (NIMH), Lauren Baskir, Ph.D.
(NIMH), Laurence Stanford, PhD. (NIDA), Karen Sirocco, Ph.D.
(NIDA) and from NINDS, Katrina Gwinn-Hardy, M.D., and Giovanna
Spinella, M.D. The Principal Investigator of the Spectroscopy Processing Center at the University of California Los Angeles is
James T. McCracken, M.D., Investigators Jeffry R. Alger, Ph.D., Jennifer Levitt, M.D., Joseph O'Neill, Ph.D.
References
Achard, S., Bullmore, E., 2007. Efficiency and cost of economical brain functional
networks. PLoS Comput. Biol. 3, e17.
Achard, S., Salvador, R., Whitcher, B., Suckling, J., Bullmore, E., 2006. A resilient,
low-frequency, small-world human brain functional network with highly
connected association cortical hubs. J. Neurosci. 26, 63–72.
Alexander-Bloch, A., Giedd, J.N., Bullmore, E., 2013. Imaging structural co-variance
between human brain regions. Nat. Rev. Neurosci. 14, 322–336.
Anderson, B.J., Alcantara, A.A., Greenough, W.T., 1996. Motor-skill learning: changes
in synaptic organization of the rat cerebellar cortex. Neurobiol. Learn. Mem. 66,
221–229.
Anderson, B.J., Li, X.C., Alcantara, A.A., Isaacs, K.R., Black, J.E., Greenough, W.T., 1994.
Glial hypertrophy is associated with synaptogenesis following motor-skill
learning, but not with angiogenesis following exercise. Glia 11, 73–80.
Andreasen, N.C., Flaum, M., Swayze 2nd, V., O'Leary, D.S., Alliger, R., Cohen, G.,
Ehrhardt, J., Yuh, W.T., 1993. Intelligence and brain structure in normal individuals. Am. J. Psychiatry 150, 130–134.
Bassett, D.S., Bullmore, E., Verchinski, B.A., Mattay, V.S., Weinberger, D.R., MeyerLindenberg, A., 2008. Hierarchical organization of human cortical networks in
health and schizophrenia. J. Neurosci. 28, 9239–9248.
Bassett, D.S., Bullmore, E.T., 2009. Human brain networks in health and disease.
Curr. Opin. Neurol. 22, 340–347.
Bermudez, P., Lerch, J.P., Evans, A.C., Zatorre, R.J., 2009. Neuroanatomical correlates
of musicianship as revealed by cortical thickness and voxel-based morphometry. Cereb. Cortex 19, 1583–1596.
Bishop, D.V., 2013. Cerebral asymmetry and language development: cause, correlate, or consequence? Science 340, 1230531.
Black, J.E., Isaacs, K.R., Anderson, B.J., Alcantara, A.A., Greenough, W.T., 1990.
Learning causes synaptogenesis, whereas motor activity causes angiogenesis, in
cerebellar cortex of adult rats. Proc. Natl. Acad. Sci. USA 87, 5568–5572.
Boivin, M.J., Giordani, B., Berent, S., Amato, D.A., Lehtinen, S., Koeppe, R.A., Buchtel,
H.A., Foster, N.L., Kuhl, D.E., 1992. Verbal fluency and positron emission tomographic mapping of regional cerebral glucose-metabolism. Cortex 28,
231–239.
Broca, M.P., 1861. Remarques sur le sie`ge de la faculte´ du langage articule´, suivies
d’une observation d’aphemie (Perte de la Parole). Bull. Soc. Anat. Paris 36,
330–357.
Bullmore, E., Sporns, O., 2009. Complex brain networks: graph theoretical analysis
of structural and functional systems. Nat. Rev. Neurosci. 10, 186–198.
Bullmore, E.T., Suckling, J., Overmeyer, S., Rabe-Hesketh, S., Taylor, E., Brammer, M.
J., 1999. Global, voxel, and cluster tests, by theory and permutation, for a difference between two groups of structural MR images of the brain. IEEE Trans.
Med. Imaging 18, 32–42.
Burgaleta, M., Johnson, W., Waber, D.P., Colom, R., Karama, S., 2014. Cognitive ability
changes and dynamics of cortical thickness development in healthy children
and adolescents. NeuroImage 84, 810–819.
Cattell, R.B., 1943. The measurement of adult intelligence. Psychol. Bull. 40,
153–193.
Chen, Z.J., He, Y., Rosa-Neto, P., Germann, J., Evans, A.C., 2008. Revealing modular
architecture of human brain structural networks by using cortical thickness
from MRI. Cereb. Cortex 18, 2374–2381.
Chklovskii, D.B., Mel, B.W., Svoboda, K., 2004. Cortical rewiring and information
storage. Nature 431, 782–788.
Choi, Y.Y., Shamosh, N.A., Cho, S.H., DeYoung, C.G., Lee, M.J., Lee, J.M., Kim, S.I., Cho,
Z.H., Kim, K., Gray, J.R., Lee, K.H., 2008. Multiple bases of human intelligence
revealed by cortical thickness and neural activation. J. Neurosci. 28,
10323–10329.
Collins, D.L., Neelin, P., Peters, T.M., Evans, A.C., 1994. Automatic 3D intersubject
registration of MR volumetric data in standardized Talairach space. J. Comput.
Assist. Tomogr. 18, 192–205.
Colom, R., Burgaleta, M., Roman, F.J., Karama, S., Alvarez-Linera, J., Abad, F.J., Martinez, K., Quiroga, M.A., Haier, R.J., 2013. Neuroanatomic overlap between intelligence and cognitive factors: morphometry methods provide support for the
key role of the frontal lobes. NeuroImage 72, 143–152.
Colom, R., Haier, R.J., Head, K., Álvarez-Linera, J., Quiroga, M.Á., Shih, P.C., Jung, R.E.,
2009. Gray matter correlates of fluid, crystallized, and spatial intelligence:
testing the P-FIT model. Intelligence 37, 124–135.
Cohen, J., Cohen, P., 1983. Applied Multiple Regression/Correlation Analysis for the
behavioral Sciences. Erlbaum, Hillsdale, NJ.
Diamond, M.C., 2001. Response of the brain to enrichment. Acad. Bras. Cienc. 73,
211–220.
Dronkers, N.F., Plaisant, O., Iba-Zizen, M.T., Cabanis, E.A., 2007. Paul Broca's historic
cases: high resolution MR imaging of the brains of Leborgne and Lelong. Brain
130, 1432–1441.
Dronkers, N.F., Wilkins, D.P., Van Valin Jr., R.D., Redfern, B.B., Jaeger, J.J., 2004. Lesion
analysis of the brain areas involved in language comprehension. Cognition 92,
145–177.
Duncan, J., Emslie, H., Williams, P., Johnson, R., Freer, C., 1996. Intelligence and the
frontal lobe: the organization of goal-directed behavior. Cogn. Psychol. 30,
257–303.
Duncan, J., Owen, A.M., 2000. Common regions of the human frontal lobe recruited
by diverse cognitive demands. Trends Neurosci. 23, 475–483.
Duncan, J., Seitz, R.J., Kolodny, J., Bor, D., Herzog, H., Ahmed, A., Newell, F.N., Emslie,
H., 2000. A neural basis for general intelligence. Science 289, 457–460.
Evans, A.C., 2013. Networks of anatomical covariance. NeuroImage 80, 489–504.
Evans, A.C., Brain Development Cooperative, G., 2006. The NIH MRI study of normal
brain development. NeuroImage 30, 184–202.
Fornito, A., Bullmore, E.T., 2012. Connectomic intermediate phenotypes for psychiatric disorders. Front. Psychiatry 3, 32.
Fornito, A., Zalesky, A., Breakspear, M., 2015. The connectomics of brain disorders.
Nat. Rev. Neurosci. 16, 159–172.
Genovese, C.R., Lazar, N.A., Nichols, T., 2002. Thresholding of statistical maps in
functional neuroimaging using the false discovery rate. NeuroImage 15,
870–878.
Gong, G., He, Y., Chen, Z.J., Evans, A.C., 2012. Convergence and divergence of
thickness correlations with diffusion connections across the human cerebral
cortex. NeuroImage 59, 1239–1248.
Gong, G., Rosa-Neto, P., Carbonell, F., Chen, Z.J., He, Y., Evans, A.C., 2009. Age- and
gender-related differences in the cortical anatomical network. J. Neurosci. 29,
15684–15693.
Gotts, S.J., Jo, H.J., Wallace, G.L., Saad, Z.S., Cox, R.W., Martin, A., 2013. Two distinct
forms of functional lateralization in the human brain. Proc. Natl. Acad. Sci. USA
110, E3435–E3444.
Gray, J.R., Chabris, C.F., Braver, T.S., 2003. Neural mechanisms of general fluid intelligence. Nat. Neurosci. 6, 316–322.
Guimera, R., Mossa, S., Turtschi, A., Amaral, L.A.N., 2005. The worldwide air transportation network: anomalous centrality, community structure, and cities'
global roles. Proc. Natl. Acad. Sci. USA 102, 7794–7799.
Haier, R.J., Siegel, B.V., Nuechterlein, K.H., Hazlett, E., Wu, J.C., Paek, J., Browning, H.
L., Buchsbaum, M.S., 1988. Cortical glucose metabolic-rate correlates of abstract
reasoning and attention studied with positron emission tomography.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
Intelligence 12, 199–217.
He, Y., Chen, Z., Gong, G., Evans, A., 2009. Neuronal networks in Alzheimer's disease.
Neuroscientist 15, 333–350.
He, Y., Chen, Z.J., Evans, A.C., 2007. Small-world anatomical networks in the human
brain revealed by cortical thickness from MRI. Cereb. Cortex 17, 2407–2419.
He, Y., Evans, A., 2010. Graph theoretical modeling of brain connectivity. Curr. Opin.
Neurol. 23, 341–350.
He, Y., Chen, Z., Evans, A., 2008. Structural insights into aberrant topological patterns of large-scale cortical networks in Alzheimer's Disease. J. Neurosci. 28,
4756–4766.
He, Y., Dagher, A., Chen, Z., Charil, A., Zijdenbos, A., Worsley, K., Evans, A., 2009a.
Impaired small-world efficiency in structural cortical networks in multiple
sclerosis associated with white matter lesion load. Brain 132, 3366–3379.
He, Y., Wang, J., Wang, L., Chen, Z.J., Yan, C., Yang, H., Tang, H., Zhu, C., Gong, Q., Zang,
Y., Evans, A.C., 2009b. Uncovering intrinsic modular organization of spontaneous brain activity in humans. PLoS One 4, e5226.
Hickok, G., Poeppel, D., 2007. Opinion – the cortical organization of speech processing. Nat. Rev. Neurosci. 8, 393–402.
Hong, S.B., Zalesky, A., Fornito, A., Park, S., Yang, Y.H., Park, M.H., Song, I.C., Sohn, C.
H., Shin, M.S., Kim, B.N., Cho, S.C., Han, D.H., Cheong, J.H., Kim, J.W., 2014.
Connectomic disturbances in attention-deficit/hyperactivity disorder: a wholebrain tractography analysis. Biol. Psychiatry 76, 656–663.
Huttenlocher, P.R., 1991. Dendritic and synaptic pathology in mental retardation.
Pediatr. Neurol. 7, 79–85.
Isaacs, K.R., Anderson, B.J., Alcantara, A.A., Black, J.E., Greenough, W.T., 1992. Exercise and the brain: angiogenesis in the adult rat cerebellum after vigorous
physical activity and motor skill learning. J. Cereb. Blood Flow Metab. 12,
110–119.
Jacobs, B., Scheibel, A.B., 1993. A quantitative dendritic analysis of Wernicke's area
in humans. I. Lifespan changes. J. Comp. Neurol. 327, 83–96.
Kabani, N., Le Goualher, G., MacDonald, D., Evans, A.C., 2001. Measurement of
cortical thickness using an automated 3-D algorithm: a validation study. NeuroImage 13, 375–380.
Kane, M.J., Engle, R.W., 2002. The role of prefrontal cortex in working-memory
capacity, executive attention, and general fluid intelligence: an individual-differences perspective. Psychon. Bull. Rev. 9, 637–671.
Karama, S., Bastin, M.E., Murray, C., Royle, N.A., Penke, L., Munoz Maniega, S., Gow,
A.J., Corley, J., Valdes Hernandez Mdel, C., Lewis, J.D., Rousseau, M.E., Lepage, C.,
Fonov, V., Collins, D.L., Booth, T., Rioux, P., Sherif, T., Adalat, R., Starr, J.M., Evans,
A.C., Wardlaw, J.M., Deary, I.J., 2014. Childhood cognitive ability accounts for
associations between cognitive ability and brain cortical thickness in old age.
Mol. Psychiatry 19, 555–559.
Karama, S., Colom, R., Johnson, W., Deary, I.J., Haier, R., Waber, D.P., Lepage, C.,
Ganjavi, H., Jung, R., Evans, A.C., Brain Development Cooperative, G., 2011.
Cortical thickness correlates of specific cognitive performance accounted for by
the general factor of intelligence in healthy children aged 6 to 18. NeuroImage
55, 1443–1453.
Kaufman, A.S., Kaufman, N.L., 1983. Kaufman Assessment Battery for Children.
American Guidance Service, Circle Pines, MN.
Kelly, C., Toro, R., Di Martino, A., Cox, C.L., Bellec, P., Castellanos, F.X., Milham, M.P.,
2012. A convergent functional architecture of the insula emerges across imaging modalities. NeuroImage 61, 1129–1142.
Khundrakpam, B.S., Reid, A., Brauer, J., Carbonell, F., Lewis, J., Ameis, S., Karama, S.,
Lee, J., Chen, Z., Das, S., Evans, A.C., Brain Development Cooperative, G., 2013.
Developmental changes in organization of structural brain networks. Cereb.
Cortex 23, 2072–2085.
Kim, J.S., Singh, V., Lee, J.K., Lerch, J., Ad-Dab'bagh, Y., MacDonald, D., Lee, J.M., Kim,
S.I., Evans, A.C., 2005. Automated 3-D extraction and evaluation of the inner
and outer cortical surfaces using a Laplacian map and partial volume effect
classification. NeuroImage 27, 210–221.
Kleim, J.A., Lussnig, E., Schwarz, E.R., Comery, T.A., Greenough, W.T., 1996. Synaptogenesis and Fos expression in the motor cortex of the adult rat after
motor skill learning. J. Neurosci. 16, 4529–4535.
Klingberg, T., Hedehus, M., Temple, E., Salz, T., Gabrieli, J.D., Moseley, M.E., Poldrack,
R.A., 2000. Microstructure of temporo-parietal white matter as a basis for
reading ability: evidence from diffusion tensor magnetic resonance imaging.
Neuron 25, 493–500.
Kyllonen, P.C., Christal, R.E., 1990. Reasoning ability is (little more than) workingmemory capacity. Intelligence 14, 389–433.
la Fougere, C., Grant, S., Kostikov, A., Schirrmacher, R., Gravel, P., Schipper, H.M.,
Reader, A., Evans, A., Thiel, A., 2011. Where in-vivo imaging meets cytoarchitectonics: the relationship between cortical thickness and neuronal density
measured with high-resolution 18F flumazenil-PET. NeuroImage 56, 951–960.
Lackner, J.R., Teuber, H.L., 1973. Alterations in auditory fusion thresholds after
cerebral injury in man. Neuropsychologia 11, 409–415.
Langer, N., Pedroni, A., Gianotti, L.R., Hanggi, J., Knoch, D., Jancke, L., 2012. Functional brain network efficiency predicts intelligence. Hum. Brain Mapp. 33,
1393–1406.
Latora, V., Marchiori, M., 2001. Efficient behavior of small-world networks. Phys.
Rev. Lett. 87, 198701.
Lee, H., Devlin, J.T., Shakeshaft, C., Stewart, L.H., Brennan, A., Glensman, J., Pitcher,
K., Crinion, J., Mechelli, A., Frackowiak, R.S., Green, D.W., Price, C.J., 2007.
Anatomical traces of vocabulary acquisition in the adolescent brain. J. Neurosci.
27, 1184–1189.
Lee, N.R., Raznahan, A., Wallace, G.L., Alexander-Bloch, A., Clasen, L.S., Lerch, J.P.,
Giedd, J.N., 2014. Anatomical coupling among distributed cortical regions in
13
youth varies as a function of individual differences in vocabulary abilities. Hum.
Brain Mapp. 35, 1885–1895.
Lerch, J.P., Evans, A.C., 2005. Cortical thickness analysis examined through power
analysis and a population simulation. NeuroImage 24, 163–173.
Lerch, J.P., Pruessner, J.C., Zijdenbos, A., Hampel, H., Teipel, S.J., Evans, A.C., 2005.
Focal decline of cortical thickness in Alzheimer's disease identified by computational neuroanatomy. Cereb. Cortex 15, 995–1001.
Lerch, J.P., Worsley, K., Shaw, W.P., Greenstein, D.K., Lenroot, R.K., Giedd, J., Evans, A.
C., 2006. Mapping anatomical correlations across cerebral cortex (MACACC)
using cortical thickness from MRI. NeuroImage 31, 993–1003.
Lerch, J.P., Yiu, A.P., Martinez-Canabal, A., Pekar, T., Bohbot, V.D., Frankland, P.W.,
Henkelman, R.M., Josselyn, S.A., Sled, J.G., 2011. Maze training in mice induces
MRI-detectable brain shape changes specific to the type of learning. NeuroImage 54, 2086–2095.
Li, Y., Liu, Y., Li, J., Qin, W., Li, K., Yu, C., Jiang, T., 2009. Brain anatomical network and
intelligence. PLoS Comput. Biol. 5, e1000395.
MacDonald, D., Kabani, N., Avis, D., Evans, A.C., 2000. Automated 3-D extraction of
inner and outer surfaces of cerebral cortex from MRI. NeuroImage 12, 340–356.
MacLullich, A.M., Ferguson, K.J., Deary, I.J., Seckl, J.R., Starr, J.M., Wardlaw, J.M.,
2002. Intracranial capacity and brain volumes are associated with cognition in
healthy elderly men. Neurology 59, 169–174.
Maxwell, A.E., Fenwick, P.B., Fenton, G.W., Dollimore, J., 1974. Reading ability and
brain function: a simple statistical model. Psychol. Med. 4, 274–280.
Mechelli, A., Crinion, J.T., Noppeney, U., O'Doherty, J., Ashburner, J., Frackowiak, R.S.,
Price, C.J., 2004. Neurolinguistics: structural plasticity in the bilingual brain.
Nature 431, 757.
Naglieri, J.A., Bornstein, B.T., 2003. Intelligence and achievement: just how correlated are they? J. Pyschoeduc. Assess. 21, 244–260.
Naglieri, J.A., Das, J.P., 1997. Cognitive Assessment System. Administration and
scoring Manual. Interpretive Handbook. Riverside, Itasca, IL.
Narr, K.L., Woods, R.P., Thompson, P.M., Szeszko, P., Robinson, D., Dimtcheva, T.,
Gurbani, M., Toga, A.W., Bilder, R.M., 2007. Relationships between IQ and regional cortical gray matter thickness in healthy adults. Cereb. Cortex 17,
2163–2171.
Neubauer, A.C., Fink, A., 2009. Intelligence and neural efficiency: measures of brain
activation versus measures of functional connectivity in the brain. Intelligence
37, 223–229.
Newman, M.E., 2006. Modularity and community structure in networks. Proc. Natl.
Acad. Sci. USA 103, 8577–8582.
Poeppel, D., 2003. The analysis of speech in different temporal integration windows: cerebral lateralization as 'asymmetric sampling in time'. Speech Commun. 41, 245–255.
Rakic, P., 1988. Specification of cerebral cortical areas. Science 241, 170–176.
Ramsden, S., Richardson, F.M., Josse, G., Thomas, M.S., Ellis, C., Shakeshaft, C.,
Seghier, M.L., Price, C.J., 2011. Verbal and non-verbal intelligence changes in the
teenage brain. Nature 479, 113–116.
Raznahan, A., Lerch, J.P., Lee, N., Greenstein, D., Wallace, G.L., Stockman, M., Clasen,
L., Shaw, P.W., Giedd, J.N., 2011. Patterns of coordinated anatomical change in
human cortical development: a longitudinal neuroimaging study of maturational coupling. Neuron 72, 873–884.
Reichle, E.D., Carpenter, P.A., Just, M.A., 2000. The neural bases of strategy and skill
in sentence-picture verification. Cogn. Psychol. 40, 261–295.
Richlan, F., 2012. Developmental dyslexia: dysfunction of a left hemisphere reading
network. Front. Hum. Neurosci. 6, 120.
Rubinov, M., Sporns, O., 2010. Complex network measures of brain connectivity:
uses and interpretations. NeuroImage 52, 1059–1069.
Schmithorst, V.J., Wilke, M., Dardzinski, B.J., Holland, S.K., 2005. Cognitive functions
correlate with white matter architecture in a normal pediatric population: a
diffusion tensor MRI study. Hum. Brain Mapp. 26, 139–147.
Semmes, J., 1968. Hemispheric specialization – a possible clue to mechanism.
Neuropsychologia 6, 11-&.
Sharda, M., Khundrakpam, B.S., Evans, A.C., Singh, N.C., 2014. Disruption of structural covariance networks for language in autism is modulated by verbal ability.
Brain Struct. Funct.
Shaw, P., Greenstein, D., Lerch, J., Clasen, L., Lenroot, R., Gogtay, N., Evans, A., Rapoport, J., Giedd, J., 2006. Intellectual ability and cortical development in children and adolescents. Nature 440, 676–679.
Sled, J.G., Zijdenbos, A.P., Evans, A.C., 1998. A nonparametric method for automatic
correction of intensity nonuniformity in MRI data. IEEE Trans. Med. Imaging 17,
87–97.
Sporns, O., 2013. Making sense of brain network data. Nat. Methods 10, 491–493.
Sur, M., Rubenstein, J.L., 2005. Patterning and plasticity of the cerebral cortex.
Science 310, 805–810.
Talairach, J., Tournoux, P., 1988. Co-Planar Stereotaxic Atlas of the Human Brain.
Thieme, New York.
Tzourio-Mazoyer, N., Landeau, B., Papathanassiou, D., Crivello, F., Etard, O., Delcroix,
N., Mazoyer, B., Joliot, M., 2002. Automated anatomical labeling of activations in
SPM using a macroscopic anatomical parcellation of the MNI MRI single-subject
brain. NeuroImage 15, 273–289.
van den Heuvel, M.P., Fornito, A., 2014. Brain networks in schizophrenia. Neuropsychol. Rev. 24, 32–48.
van den Heuvel, M.P., Stam, C.J., Kahn, R.S., Hulshoff Pol, H.E., 2009. Efficiency of
functional brain networks and intellectual performance. J. Neurosci. 29,
7619–7624.
van der Mark, S., Klaver, P., Bucher, K., Maurer, U., Schulz, E., Brem, S., Martin, E.,
Brandeis, D., 2011. The left occipitotemporal system in reading: disruption of
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i
14
B.S. Khundrakpam et al. / NeuroImage ∎ (∎∎∎∎) ∎∎∎–∎∎∎
focal fMRI connectivity to left inferior frontal and inferior parietal language
areas in children with dyslexia. NeuroImage 54, 2426–2436.
Waber, D.P., De Moor, C., Forbes, P.W., Almli, C.R., Botteron, K.N., Leonard, G., Milovan, D., Paus, T., Rumsey, J., Brain Development Cooperative, G., 2007. The NIH
MRI study of normal brain development: performance of a population based
sample of healthy children aged 6 to 18 years on a neuropsychological battery.
J. Int. Neuropsychol. Soc. 13, 729–746.
Waltz, J.A., Knowlton, B.J., Holyoak, K.J., Boone, K.B., Mishkin, F.S., Santos, M.D.,
Thomas, C.R., Miller, B.L., 1999. A system for relational reasoning in human
prefrontal cortex. Psychol. Sci. 10, 119–125.
Wang, D., Buckner, R.L., Liu, H., 2014. Functional specialization in the human brain
estimated by intrinsic hemispheric interaction. J. Neurosci. 34, 12341–12352.
Wechsler, D., 1999. Wechsler Abbreviated Scale of Intelligence (WASI). Psychological Corporation, San Antonio, TX.
Witelson, S.F., Beresh, H., Kigar, D.L., 2006. Intelligence and brain size in 100
postmortem brains: sex, lateralization and age factors. Brain 129, 386–398.
Zatorre, R.J., Fields, R.D., Johansen-Berg, H., 2012. Plasticity in gray and white:
neuroimaging changes in brain structure during learning. Nat. Neurosci. 15,
528–536.
Zielinski, B.A., Gennatas, E.D., Zhou, J., Seeley, W.W., 2010. Network-level structural
covariance in the developing brain. Proc. Natl. Acad. Sci. USA 107, 18191–18196.
Zijdenbos, A.P., Forghani, R., Evans, A.C., 2002. Automatic "pipeline" analysis of 3-D
MRI data for clinical trials: application to multiple sclerosis. IEEE Trans. Med.
Imaging 21, 1280–1291.
Please cite this article as: Khundrakpam, B.S., et al., Imaging structural covariance in the development of intelligence. NeuroImage
(2016), http://dx.doi.org/10.1016/j.neuroimage.2016.08.041i