brain research 1558 (2014) 33–43

Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report

Functional organization of intrinsic connectivity networks in Chinese-chess experts Xujun Duana,1, Zhiliang Longa,1, Huafu Chena,n, Dongmei Liangb,c, Lihua Qiud, Xiaoqi Huangd, Timon Cheng-Yi Liub,n, Qiyong Gongd a Key Laboratory for NeuroInformation of Ministry of Education, School of Life Science and Technology, University of Electronic Science and Technology of China, Chengdu 610054, PR China b School of Physical Education & Sports Exercise, South China Normal University, University Town, Guangzhou 510006, PR China c Center for Studies of Psychological Application, School of Psychology, South China Normal University, Guangzhou 510631, PR China d Huaxi MR Research Center, Department of Radiology, West China Hospital of Sichuan University, 610041, PR China

art i cle i nfo

ab st rac t

Article history:

The functional architecture of the human brain has been extensively described in terms of

Accepted 17 February 2014

functional connectivity networks, detected from the low-frequency coherent neuronal

Available online 22 February 2014

fluctuations during a resting state condition. Accumulating evidence suggests that the

Keywords:

overall organization of functional connectivity networks is associated with individual

Chess expert

differences in cognitive performance and prior experience. Such an association raises the

Resting-state fMRI

question of how cognitive expertise exerts an influence on the topological properties of

Functional connectivity network

large-scale functional networks. To address this question, we examined the overall

Graph theoretical analysis

organization of brain functional networks in 20 grandmaster and master level Chinese-

Small-world topology

chess players (GM/M) and twenty novice players, by means of resting-state functional connectivity and graph theoretical analyses. We found that, relative to novices, functional connectivity was increased in GM/Ms between basal ganglia, thalamus, hippocampus, and several parietal and temporal areas, suggesting the influence of cognitive expertise on intrinsic connectivity networks associated with learning and memory. Furthermore, we observed economical small-world topology in the whole-brain functional connectivity networks in both groups, but GM/Ms exhibited significantly increased values of normalized clustering coefficient which resulted in increased small-world topology. These findings suggest an association between the functional organization of brain networks and individual differences in cognitive expertise, which might provide further evidence of the mechanisms underlying expert behavior. & 2014 Elsevier B.V. All rights reserved.

n

Corresponding authors. E-mail addresses: [email protected] (H. Chen), [email protected] (T.-Y. Liu). 1 Xujun Duan and Zhiliang Long contributed equally to this work.

http://dx.doi.org/10.1016/j.brainres.2014.02.033 0006-8993 & 2014 Elsevier B.V. All rights reserved.

34

1.

brain research 1558 (2014) 33–43

Introduction

Task-free spontaneous neural activity has been proposed to play an important part in maintaining ongoing representations of conscious activity in the resting brain, and it demonstrates temporal coherence between brain regions that are anatomically connected or functionally related (Biswal et al., 1995; Raichle, 2010; Zhang and Raichle, 2010). Recently, assessments of brain intrinsic functional connectivity were conducted in order to investigate the level of integration of brain systems at a resting state when no task is being performed (Greicius et al., 2003). Along with intrinsic functional connectivity analyses, graph theoretical approaches have also emerged as powerful tools to explore brain organization on the level of large-scale functional networks (Bullmore and Sporns, 2009; He and Evans, 2010). It is well accepted that functional connections of brain networks are organized in a highly efficient small-world manner, characterized by a high level of neighborhood clustering and a short average distance of nodes within the overall network (Achard et al., 2006; Sporns et al., 2004). Small-world attributes have been found in brain functional networks measured from electroencephalography, magnetoencephalography, functional magnetic resonance imaging (Bassett and Bullmore, 2006; Ferri et al., 2007; Liao et al., 2011a; Smit et al., 2008; Supekar et al., 2009), as well as in brain anatomical networks using structural magnetic resonance imaging and diffusion tensor imaging (Fan et al., 2011; Gong et al., 2009; Hagmann et al., 2007; He et al., 2007). There is increasing evidence that the small-world organization of brain networks can be affected by normal aging (Achard and Bullmore, 2007) and brain diseases, such as schizophrenia (Liu et al., 2008; Wang et al., 2012; Yu et al., 2011), Alzheimer's disease (Sanz-Arigita et al., 2010; Stam et al., 2007), attention-deficit/hyperactivity disorder (Wang et al., 2009b), and epilepsy (Liao et al., 2010; Ponten et al., 2007; Wang et al., 2010). However, it has also been suggested that learning and training can enhance functional organization on the level of large-scale brain networks. For instance, Voss et al. (2010) found that 1-year intervention trial of aerobic training improved the aging brain's resting functional connectivity as well as network efficiency in higher-level cognitive networks, providing important evidence for exercise-induced functional plasticity in large-scale brain systems. A recent study conducted by Albert et al. (2009) showed that motor learning modulated resting state networks in a positive manner, and another study performed by Lewis et al. (2009) demonstrated that visual perceptual learning modified the interaction and organization between functional networks. Moreover, several recent studies also suggested that the level of global communication efficiency of the brain network is positively associated with individual differences in cognitive performance. For instance, van den Heuvel et al. (2009) demonstrated a strong negative relationship between the normalized characteristic path length of the resting-state brain network and intelligence quotient, suggesting a positive association between the global efficiency of functional brain networks and intellectual performance. Another study performed by Li et al. (2009) showed that

people with higher intelligence tend to have greater global efficiency of the brain anatomical network. This evidence suggests an association between large-scale organization of brain networks and individual differences in cognitive performance as well as prior experiences. Such an association raises the question of how cognitive expertise exerts an influence on the topological properties of functional networks. The board game chess involves many aspects of high level cognition and requires sophisticated problem solving skills (Atherton et al., 2003), thus providing a good opportunity to study the mechanisms underlying cognitive expertise (Wan et al., 2011). In early psychological studies, researchers found that, compared with novices, world-class grandmasters searched much deeper and wider, using more efficient search processes and more complex evaluation functions to assess their decisions when detecting the best moves (Charness, 1981; Reynolds, 1982). Relative to local-club players, grandmasters tended to search at similar depth or width, but generated moves faster, reached a decision faster, and the best next-move was always generated in the very beginning of their search (Connors et al., 2011; De Groot, 1946; Gobet and Charness, 2006; Lassiter, 2000). Several brain imaging techniques have been employed to study chess skills. In general, these studies indicated that frontal and posterior parietal areas, which are known to be involved in top-down orienting of attention, perception and working memory, are engaged in chess playing (Amidzic et al., 2001; Atherton et al., 2003; Bilalic et al., 2010; Campitelli et al., 2005; Gobet and Charness, 2006; Nichelli et al., 1994; Onofrj et al., 1995). There is also some evidence demonstrating different brain activation patterns between experts and amateurs (or novices) while performing tasks related to chess. For instance, Amidzic et al. (2001) found that, compared to amateur players, highly skilled chess grandmasters had more bursts of gamma band activity in the brain regions associated with expert memory retrieval during matches. In more recent research, Wan et al. (2011) studied the neural basis of intuitive best next-move generation in Japanese chess experts, and indicated that the superior capability of board game experts largely depends on quick automatic processing skills which are mediated by the caudate nucleus. In our previous studies we studied the morphological differences of the caudate nucleus between chess experts and novices due to the important role of this region in chess skills. We found that the caudate nuclei of chess experts were significantly smaller relative to those of novice controls but exhibit increased connections with widely distributed brain regions in spontaneous oscillatory activity (Duan et al., 2012a, 2012b). These findings suggest that chess grandmasters may differ from novices in brain functional organization of both local connections and global topologies. However, almost nothing is currently known about the influence of high-level cognitive expertise on large-scale intrinsic functional connectivity, not to mention the global topological properties of the brain networks. In the current study, we hypothesize that superior chess experts might differ with novices on: (1) the functional connectivity of intrinsic brain networks, especially connections associated with learning and memory systems such as basal ganglia and medial temporal lobe, due to the important roles of these

brain research 1558 (2014) 33–43

regions in chess expertise; (2) the global topological properties of the whole-brain functional networks, according to prior evidence that learning and practice can enhance the functional organization on the level of large-scale brain networks. To test this hypothesis, we used resting-state functional magnetic resonance imaging (rs-fMRI) to construct brain functional networks of a group of grandmaster and master level Chinese-chess players (GM/M) and a group of novice players. The whole brain was divided into 90 cortical and subcortical regions, and functional connectivity was estimated between the mean time series of each pair of brain regions according to temporal correlation. The differences in wholebrain functional connectivity between GM/Ms and novices were assessed by network-based statistic (NBS) approaches. Furthermore, the whole-brain interregional correlation matrices were thresholded to generate a set of undirected weighted graphs to construct brain functional networks. Finally, the topological properties of the networks were characterized by graph theoretical approaches, and the differences between these two groups were further statistically evaluated.

2.

Results

2.1.

Increased functional connectivity in GM/Ms

We firstly employed one sample t test to identify functional connections significantly existed between pairs of nodes in

35

GM/Ms and novices. For a Bonferroni correction of po0.001, 1350 connections survived in novices, while 2184 connections survived in GM/Ms. To further localize specific pairs of brain regions in which the functional connectivity was altered in GM/Ms, we used a network based statistic (NBS) approach proposed by Zalesky et al. (2010). The NBS identified a significantly enhanced subnetwork of connections in GM/Ms comparing with novices, primarily comprising the basal ganglia, thalamus, medial temporal lobe, and several temporal and parietal areas (p¼ 0.0259, permutation testing; Fig. 1, Table 1). No significantly decreased functional connectivity was found in GM/Ms compared to novices. Results were visualized with the BrainNet Viewer (http://www.nitrc.org/projects/bnv/).

2.2.

Enhanced small-world organization in GM/Ms

The functional connectivity networks corresponding to both groups represented a clear small-world organization for a defined range of sparsity (S) thresholds, expressed by γ41 and λ  1. The evaluation of the integrated AUC values revealed that GM/Ms had significantly higher values of small-worldness index (s) and normalized weighted clustering coefficient (γ) relative to novices (permutation testing, po0.05). Since there were no significant differences in normalized weighted characteristic path length (λ), we assume that the increased s were attributable to the increased γ.

Fig. 1 – Statistically significant increased functional connectivity strength in GM/Ms compared to novice controls. Nodes are located based on their centroid stereotaxic coordinates and color coded according to the six anatomical subregions listed in Table 2, and further mapped onto the cortical surfaces at five views: lateral view of left hemisphere (upper left), lateral view of right hemisphere (upper right), medial view of left hemisphere (lower left), medial view of right hemisphere (lower right), and superior view (middle). Undirected edges have different thickness according to significnace of functional connectivity differences.

36

brain research 1558 (2014) 33–43

Table 1 – Increased functional connectivity strength in GM/Ms compared with novices revealed by NBS. Region 1 Abbreviation

Region 2 Category

Increased connections in network 1, p ¼ 0.0259, corrected PUT.L Subcortical PUT.R Subcortical PAL.R Subcortical HIP.L Medial temporal HIP.L Medial temporal HIP.L Medial temporal HIP.R Medial temporal HIP.R Medial temporal HIP.R Medial temporal HIP.R Medial temporal THA.L Subcortical THA.L Subcortical THA.L Subcortical THA.L Subcortical THA.L Subcortical

Abbreviation

Category

THA.L HIP.R HIP.R ROL.L ROL.R MTG.R ROL.R PHIP.R FG.L SMG.R ROL.L ROL.R INS.L FG.L SMG.L

Subcortical Medial temporal Medial temporal Parietal-(pre)motor Parietal-(pre)motor Temporal Parietal-(pre)motor Medial temporal Occipital Parietal-(pre)motor Parietal-(pre)motor Parietal-(pre)motor Temporal Occipital Parietal-(pre)motor

Besides λ, there were no significant differences in the AUC values of Sw , Eglobal and Elocal between the two groups (Fig. 2).

3.

Discussion

This is the first study, to our knowledge, to investigate the effect of high-level cognitive expertise in whole brain functional networks and their topological characteristics. We found that functional connectivity was increased in GM/Ms between basal ganglia, thalamus, medial temporal regions and several parietal and temporal areas, relative to novices. We further found that whole brain functional networks exhibited economical small-world topology in both groups, but GM/Ms exhibited significantly increased values of smallworld topology which might be attributable to the increased normalized clustering coefficient. These findings suggest that cognitive expertise has a positive influence on the functional sub-networks center around the brain regions associated with the expertise as well as the global topology of the whole-brain network, which might in turn facilitate withinnetwork communication for expert behavior.

3.1.

t

Enhanced functional connectivity in GM/Ms

In the present study, comparison of the functional connectivity networks between the two groups revealed that GM/Ms produced significantly stronger connectivity than the novices mostly between the basal ganglia, thalamus, medial temporal regions and several parietal and temporal areas (Fig. 1, Table 1). It has been suggested that the basal ganglia, thalamus, medial temporal regions and their anatomically adjacent cortical and subcortical structures constitute the limbic system, which plays an important role in learning and memory. Specifically, the thalamus is a critical component of the frontal cortical-basal ganglia-thalamic circuits that mediate motivation, planning and cognition for the development and expression of goal-directed behaviors (Haber and Calzavara, 2009). Notably, the thalamus sends a massive

3.69 3.32 3.51 3.36 3.53 3.39 3.38 3.34 3.42 3.35 4.13 3.57 3.41 3.53 3.68

projection directly to the basal ganglia, which is considered to be particularly critical for learning of new complex behaviors (Graybiel, 2005; Pasupathy and Miller, 2005). The basal ganglia itself is associated with reinforcement-based trial and error learning, as well as the formation and execution of habit, or the so-called stimulus-response association (Graybiel, 2005; Yin et al., 2008). Specific activation of the basal ganglia in professional chess players was detected during intuitive generation of the best next-move, consistent with the important role of basal ganglia in quick automaticprocessing cognitive skills (Wan et al., 2011). In chess experts, chunks (i.e. the units of perception in chess patterns) are stored in long-term memory in an accessible form, which give access to information with respect to what move to play and what plan to follow (Chase and Simon, 1973; Ericsson and Kintsch, 1995; Gobet and Simon, 1998). Thus perception of chunks automatically evokes the idea of the best next-move or best series of moves (de Groot and Gobet, 1996; Wan et al., 2011). This idea is similar to the formation of stimulusresponse associations, which might also be related to the strong reciprocal connections between the basal ganglia and medial temporal lobe (Dickerson et al., 2011; Voermans et al., 2004). Previous studies have confirmed that the basal ganglia and medial temporal lobe form interacting memory systems, which contribute to skill and knowledge acquisition as well as habit formation, especially during the early phases of learning, relying on both procedural and declarative memory systems (Poldrack et al., 2001; Poldrack and Rodriguez, 2004). Previous brain imaging studies have investigated the taskevoked brain activation differences between chess experts and novices (or amateurs), and demonstrated substantial differences in brain activation in the striatum, medial temporal lobe and other related regions in response to both domain-specific and domain-general tasks associated with memory, decision making, perception and problem-solving (Bilalic et al., 2011; Bukach et al., 2006; Campitelli et al., 2005, 2007; Wan et al., 2011). Those findings are of critical importance to the neural basis of chess expertise but also raise another question on whether those activation differences are

brain research 1558 (2014) 33–43

37

Fig. 2 – Between-group differences in overall topological characteristics of functional connectivity networks. The barplot indicates the statistical differences of the area under the curve (AUC) over a range of thresholds of 0.1oSo0.48 between groups. The asterisk (*) indicates a significance level of po0.05 (permutation testing). Notes: r, small-world topology; γ, normalized clustering coefficient; λ, normalized characteristic shortest path length; Sw, total connection strength; Elocal, local efficiency; Eglobal, global efficiency.

limited to specific task paradigms, or do they reflect a more fundamental differences on the functional architecture represented by spontaneous activity? In the present study, we found that, compared with novices, GM/Ms produced significantly stronger functional connectivity between the basal ganglia, thalamus, medial temporal regions and several parietal and temporal areas, reflecting that long-term and intensive practice of chess expertise modulates spontaneous neural activity in circuitries associated with learning and memory. Accumulating evidence suggests that spontaneous activity patterns recorded at rest reflects the fundamental functionalanatomic organization of the brain, which might be associated with individual differences in cognitive performance and prior experience (Fox et al., 2007; Greicius et al., 2003; Lewis et al., 2009; Mennes et al., 2010). Recently, functional connectivity methods based on spontaneous activity have emerged as powerful tools to investigate the level of integration of brain systems at a resting state when no task was performed (Greicius et al., 2003). Task-free spontaneous neural activity has been proposed to play an important part in maintaining the ongoing representations of conscious activity in the resting brain, and it demonstrates temporal coherence between brain regions that are anatomically connected or functionally related through co-activation in response to task performance (Biswal et al., 1995; Raichle,

2010; Zhang and Raichle, 2010). Previous evidence from resting-state studies has consistently demonstrated that prior experience (i.e. the history of network activation) changes resting functional connectivity in a behaviorally specific manner (Hampson et al., 2006a, 2006b; Lewis et al., 2009). For instance, Hampson et al. (2006b) demonstrated that functional connectivity of the reading circuit co-varied with individual differences in reading abilities. In another study, they investigated the functional connectivity between key nodes of the default mode network both during a working memory task and at rest, and found that performance on the working memory task was positively correlated with the strength of these functional connections not only during the working memory task, but also at rest (Hampson et al., 2006a). Therefore, increased intrinsic connectivity in the functional circuits related to the basal ganglia, thalamus, and medial temporal regions might reflect the long-term and frequent engagement of these circuits in chess practice and skill learning, and the enhanced connection in turn facilitates the communication within the network. Although chess playing involves many aspects of high level cognition including attention, execution, visuo-spatial perception and working memory (Amidzic et al., 2001; Onofrj et al., 1995; van der Maas and Wagenmakers, 2005), we did not see much between-group differences of functional connectivity in brain regions involved in those cognitive

38

brain research 1558 (2014) 33–43

processes, such as the prefrontal–parietal system. One possible explanation is that, as early psychological studies suggested, high-level chess skill did not reside in differences in short-term memory capacity, but in the number of chunks held in long-term memory (Chase and Simon, 1973; Gobet and Charness, 2006). Therefore, we assume that long-term procedural and declarative memory systems which consist of basal ganglia and medial temporal lobe rather than workingmemory systems exhibited significant difference between GM/Ms and novices are of a certain amount of plausibility.

3.2.

Enhanced small-world organization in GM/Ms

Recent resting-state fMRI studies suggest that functional connectivity networks exhibit a highly efficient small-world topology, characterized by a high level of neighborhood clustering and a short average distance of nodes within the overall network (Achard et al., 2006; Achard and Bullmore, 2007; Bassett and Bullmore, 2006; van den Heuvel et al., 2008). In agreement with previous findings, we also observed the features of small-world architecture in the functional brain networks in both groups of GM/Ms and novices, supporting the standpoint that brain networks might have been evolved to maximize cost efficiency of parallel information processing, i.e., high efficiency of parallel information transfer at low cost (Achard and Bullmore, 2007; Sporns et al., 2004). Although both GM/M and novice groups had economical small-world properties as elucidated earlier, the topological properties of small-world topology and normalized clustering coefficient were significantly increased in GM/Ms. It has been suggested that the small-world structure reflects an optimal network organization, which is associated with individual differences in cognitive performance. For instance, people with short normalized characteristic path length of the resting-state brain network tend to have higher intelligence quotient scores, suggesting a positive association between the global efficiency of functional brain networks and intellectual performance (van den Heuvel et al., 2009). Similar results have been reported for brain anatomical networks constructed from diffusion tensor imaging (Li et al., 2009). It has also been suggested that learning and training can enhance topological organization of brain networks. For instance, Voss et al. (2010) found that aerobic training improved the aging brain's resting functional connectivity as well as network efficiency in higher-level cognitive networks. In the present study, the normalized clustering coefficient (γ) showed significantly higher values in GM/Ms, implying relatively intense local connectedness of the brain functional networks in GM/Ms. As revealed by functional connectivity analysis in the first section, GM/Ms produced significantly stronger connectivity mainly in the long-term procedural and declarative memory systems which consist of basal ganglia and medial temporal lobe, demonstrating the long-term and intensive involvement of those circuits in chess practice and high-level cognitive skill learning. Since the clustering coefficient quantifies the extent of local interconnectivity or cliquishness in network (Onnela et al., 2005; Watts and Strogatz, 1998), enhanced functional connectivity in the related systems might thus result in an increase of normalized clustering coefficient. Furthermore, compared

with novices, the small-world index s also showed significantly larger values in GM/Ms, which might be caused by the increased value of normalized clustering coefficient, but it also reflects the influences of high-level cognitive expertise in global topologic characteristics.

4.

Conclusions

We investigated whole brain functional connectivity networks and their topological properties in two groups of GM/ Ms and novices. Our results demonstrate an increased functional connectivity between the basal ganglia, thalamus, medial temporal regions and several parietal and temporal areas in GM/Ms compared to novices, which suggests an effect of cognitive expertise on learning and memory systems in GM/Ms. Furthermore, both groups showed a clear smallworld property on the whole-brain functional connectivity networks, but GM/Ms exhibited significantly increased values of small-world topology which might be attributable to the increased normalized clustering coefficient. Our findings highlight the influence of high-level cognitive expertise on topological organization of large-scale functional networks by improving the brain’s resting functional connectivity as well as global network characteristics, which might in turn be beneficial for specific expert behavior.

5.

Experimental procedures

5.1.

Participants

Two groups of subjects were recruited and studied. The first group consisted of twenty grandmaster and master level Chinese-chess players (GM/M) (seven female, aged 25.4576.13 years) who had a mean period of 11.877.7 years of tournament and training practice and scored between 2200 and 2600 on Elo's chess-skill rating scale (Elo, 1978). All of them were recruited from the First National Intelligence Games held in Chengdu, China. The second group consisted of twenty wellmatched novice players (seven female, aged 27.2077.84 years) who knew the rules and simple strategies of playing Chinese chess. All subjects were right-handed and had no history of psychiatric or neurological disorders. Both groups were tested by Raven's Standard Progressive Matrices, and the two groups did not differ on observation skills and clear-thinking ability according to the test (p¼ 0.63, t¼  0.44). The present study was approved by the local Institutional Review Board of the West China Hospital of Sichuan University, and informed written consent was obtained from all subjects.

5.2.

Data acquisition and preprocessing

Scanning was performed on a 3T Siemens Trio system (Erlangen, German) at the MR Research Center of West China Hospital of Sichuan University, Chengdu, China. Functional images covering the whole brain were acquired axially using a T2n-weighted gradient-echo echo-planar pulse sequence (TR, 2000 ms; TE, 30 ms; flip angle¼901; voxel size¼ 3.75  3.75  5 mm3; 5 mm thickness without gap, 30 axial slices), during

brain research 1558 (2014) 33–43

which subjects were instructed to relax with their eyes open and focus on a cross-hair centered in the screen. For each subject, the resting-state fMRI scanning lasted 410 seconds, with 205 volumes recorded in total. The first 5 volumes were discarded for magnetization stabilization, and the remaining 200 consecutive volumes were used for data analysis. Image preprocessing was performed using SPM8 software package (http://www.fil.ion.ucl. ac.uk/spm). After slice timing adjustment and realignment for head motion correction, any data affected by head motion of over 1 mm or rotation of more than 11 was excluded (no subject was excluded in the present study). We also evaluated group differences in translation and rotation of head motion according to the following formula (Liao et al., 2011b; Liu et al., 2008): Head motion rotation L qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 ¼ ∑ jxi xi  1 j2 þ jyi yi  1 j2 þ jzi  zi  1 j2 ¼ L 1 i¼2 where L is the length of the time series (L ¼ 200 in this study), xi , yi , and zi are translations/rotations at the ith time point in the x, y, and z directions, respectively. The results showed that there were no significant group differences between GM/Ms and novices in head motion (two sample two-tailed t-test, t¼ 0.642, p¼ 0.525) and rotation (two sample two-tailed t-test, t¼ 0.657, p¼0.515). The realigned images were spatially normalized into a standard stereotaxic space at 333 mm3, using the Montreal Neurological Institute (MNI) echo-planar imaging (EPI) template. In order to avoid introducing artificial local spatial correlation, no spatial smoothing was applied, as previously suggested (Achard et al., 2006; Achard and Bullmore, 2007; Salvador et al., 2005; Wang et al., 2009a).

5.3.

Functional connectivity network construction

Nodes and edges between nodes are two fundamental elements in a network. To determine the nodes of brain functional connectivity networks, we used the automated anatomical labeling (AAL) algorithm (Tzourio-Mazoyer et al., 2002) to parcellate the whole brain into 90 non-cerebellar anatomical regions of interest (45 regions of interest for each hemisphere), with each representing a node in the network. Anatomic labels of nodes are presented in Table 2. To determine the edges between pairs of nodes, we conducted the following procedures. For each subject, the representative time series in each region of interest was obtained by averaging fMRI time series across all voxels in the region of interest. Several sources of spurious variance, along with their temporal derivatives, were removed from the data through linear regression: six parameters obtained by rigid body correction of head motion, averaged signals from CSF, and averaged signals from white matter. To reduce the effect of low-frequency drift and high-frequency noise, temporal band-pass filtering (0.01–0.08 Hz) was then performed. Finally, a square N  N (where N ¼90 is the number of regions of interest) correlation matrix (i.e. functional connectivity matrix) was obtained for each subject by computing Pearson correlation coefficients between the preprocessed time series of every pair of regions of interest. Of note, both positive and negative correlation coefficients were entered into the

39

correlation matrices. A Fisher's r-to-z transformation was applied to the correlation matrices to improve the normality of the correlation coefficients (Liao et al., 2010; Liu et al., 2008).

5.4.

Network-based statistic

In this study, network-based statistic (NBS) was used to evaluate connected sub-networks consisting of enhanced functional connections in GM/Ms. Before the NBS, one sample t test was employed to test the hypothesis that functional connectivity strength between specific pairs of nodes were significantly larger than zero in both groups of GM/Ms and novices (Bai et al., 2012). Here, we used a rigorous threshold to control false positive connections by Bonferroni correction with po0.001. The resulting functional connections in GM/Ms and novices were then combined to build a connection mask, which was used for the following NBS analysis. The NBS procedure was implemented as follows: a primary threshold (here, t¼3.3) was used for each connection within the connection mask to define a set of suprathreshold functional connections evident in GM/Ms. The connected components within the set of suprathreshold connections and the component size (numbers of connections) were determined. To evaluate the significance of each connected components, we performed nonparametric permutation test (5000 permutations) to generate a null distribution of the component size. Specifically, for each permutation, we randomly assigned participants to one of the two groups with the same size as the origin groups of GM/Ms and novices. The same threshold (t¼3.3) was then applied to connections in the connection mask to determine the set of significantly increased connections in GM/Ms, and the maximum component size was computed. We then assigned a p value to each connected component by computing the proportion of null distribution values exceeding the original component size. The connected components with the p value lower than 0.05 were considered as significant. A detailed description of the NBS methodology is given by Zalesky et al. (2010).

5.5.

Graph–theory analysis

5.5.1.

Threshold selection

To investigate the topological properties of brain functional networks, each correlation matrix was thresholded into a set of weighted graphs (i.e. networks), by selecting the strongest connections until the desired sparsity was reached. In those graphs nodes represented brain regions and edges represented undirected connections. In this study, connection sparsity S was adopted as a threshold measurement. S was defined as the ratio of the existing edges to the maximum possible number of edges in a network. The sparsity threshold ensured that all resultant networks had the same number of nodes and edges, and enabled us to compare the betweengroup differences in relative network organization (Achard and Bullmore, 2007). Instead of selecting a single threshold, we applied a threshold range of 0.1oSo0.48 with an interval of 0.01 to each weighted matrix. The minimum threshold (S¼ 0.1) was computed according to the criterion that the average degree over all nodes of each thresholded network was larger than 2 log(N), where N denotes the number of

40

brain research 1558 (2014) 33–43

Table 2 – Anatomical subregions and corresponding abbreviated regional labels. Region name

Abbreviation LH

RH

Medial temporal Amygdala Hippocampus Parahippocampalgyrus Middle temporal gyrus: temporal pole Superior temporal gyrus: temporal pole

AMYG.L HIP.L PHIP.L MTGp.L STGp.L

AMYG.R HIP.R PHIP.R MPGp.R STGp.R

Frontal Anterior cingulate gyrus Gyrus rectus Inferior frontal gyrus, opercular part Inferior frontal gyrus, orbital Inferior frontal gyrus, triangular Middle frontal gyrus Middle frontal gyrus, orbital Superior frontal gyrus Superior frontal gyrus, medial Superior frontal gyrus, medial orbital Superior frontal gyrus, orbital

ACC.L REG.L IFGoper.L IFGorb.L IFGtri.L MFG.L MFGorb.L SFG.L SFGmed.L SFGmorb.L SFGorb.L

ACC.R REG.R IFGoper.R IFGorb.R IFGtri.R MFG.R MFGorb.R SFG.R SFGmed.R SFGmorb.R SFGorb.R

Occipital Calcarine fissure Cuneus Fusiform gyrus Inferior occipital gyrus Lingual gyrus Middle occipital gyrus Superior occipital gyrus

CAL.L CUN.L FG.L IOG.L LING.L MOG.L SOG.L

CAL.R CUN.R FG.R IOG.R LING.R MOG.R SOG.R

Parietal-(pre)motor Angular gyrus Inferior parietal gyrus Superior parietal gyrus Median cingulate gyri Posterior cingulate gyrus Paracentral lobule Postcentralgyrus Precentralgyrus Precuneus Supplementary motor area Supramarginalgyrus Rolandic operculum

ANG.L IPG.L SPG.L MCC.L PCC.L PCL.L PoCG.L PreCG.L PCUN.L SMA.L SMG.L ROL.L

ANG.R IPG.R SPG.R MCC.R PCC.R PCL.R PoCG.R PreCG.R PCUN.R SMA.R SMG.R ROL.R

Subcortical Caudate nucleus Olfactory cortex Pallidum Putamen Thalamus

CAU.L OLF.L PAL.L PUT.L THA.L

CAU.R OLF.R PAL.R PUT.R THA.R

Temporal Heschlgyrus Inferior temporal gyrus Insula Middle temporal gyrus Superior temporal gyrus

HES.L ITG.L INS.L MTG.L STG.L

HES.R ITG.R INS.R MTG.R STG.R

nodes. The maximum threshold (S ¼0.48) was adopted based on the criterion that the average small-worldness scalar (s, see below for definition) of thresholded network across subjects was larger than 1.1. The generated threshold range

of 0.1oSo0.48 guaranteed that the thresholded network had estimable small-worldness and control of spurious edges possibly. The following network analysis was repeatedly performed on this defined threshold range.

brain research 1558 (2014) 33–43

5.5.2.

Network characteristics

For brain networks at each sparsity threshold S, we computed global network characteristics including (1) small-world parameters involving weighted clustering coefficient Cw , weighted characteristic path length Lw , normalized weighted clustering coefficient γ, normalized weighted characteristic path length λ, and small-worldness s; (2) network efficiency involving local efficiency Elocal and global efficiency Eglobal ; and (3) total connection strength Sw . Small-world properties were originally proposed by Watts and Strogatz (1998). The Cw of a network was expressed as the average of clustering coefficients across all nodes in the network: Cw ¼ 1=N∑i A N Cw i , where N denotes the numbers of nodes, and Cw i denotes the clustering coefficient of a node i, which was defined as the likelihood that the neighborhoods of node i were connected to each other or not: Cw i ¼

1 ∑ ðw w w Þ1=3 ki ðki 1Þ j;h A N ij ih jh

where ki is the number of edges connecting to node i, and wij is the correlation weight between node i and node j. The Cw quantifies the extent of local interconnectivity or cliquishness in network (Onnela et al., 2005; Watts and Strogatz, 1998). The path length between node i and node j was defined as the sum of edge length along this path. For weighted functional connectivity networks in this study, we computed the reciprocal of edge weight, 1=wij as the length of each weighted edge. The path length Lij was defined as the length of path for node i and node j with the shortest length: w w Lw ij ¼ ∑auv A gi-j 1=wuv , where gi-j is the shortest weighted path between i and j. The weighted characteristic path length Lw of a network, which was measured by a ‘harmonic mean’ length between pairs to overcome the problem of possibly disconnected network components (Newman, 2003; Rubinov and Sporns, 2010; Stam et al., 2009): Lw ¼

1 ∑N 1=Lw 1=ðNðN 1ÞÞ∑N ij i ¼ 1 jai

A real world would be considered small world if it has similar path length but higher clustering coefficient than λ ¼ Lw = random network, which is γ ¼ Cw =Cw random 41, w w Lrandom  1 (Watts and Strogatz, 1998), where Crandom and Lw random are the mean weighted clustering coefficient and weighted characteristic path length of 100 matched random networks that preserve the same number of nodes, edges, and degree distribution as the real network. These two measures are summarized into a scalar metric, small-worldness, s ¼ γ=λ, which is typically larger than 1 in the case of small-world organization (Achard et al., 2006; Humphries et al., 2006). The global efficiency of a network, Eglobal , was defined by the inverse of the harmonic mean of the path length between each pair of nodes, which is expressed as follows: Eglobal ¼

1 1 ∑ NðN 1Þ i a j A N Lij

where N is the number of nodes in a network. The global efficiency is a measure of parallel information transformation (Achard and Bullmore, 2007).

41

The local efficiency of a network, Elocal , is defined as the average of local efficiency of each node, that is Elocal ¼

1 ∑E ðG Þ; N i A G global i

where Eglobal ðGi Þ is the global efficiency of the neighborhood subgraph Gi of the node i. The local efficiency can be understood as a measure of fault tolerance of the network, indicating how well each subgraph exchanges information when the index node is eliminated (Achard and Bullmore, 2007). The total connection strength (Sw ) of a network was computed as the average of nodal connection strength (Sw i ) for all nodes in the network: Sw ¼

1 ∑ Sw N iAN i

where N is the number of nodes in network, and Sw i was defined as the sum of the weights of all connections of node i, which is Sw i ¼ ∑j A N wij .

5.5.3.

Statistical analysis

To evaluate the between-group differences in overall topological characteristics, we calculated the area under the curve (AUC) over the whole range of thresholds of 0.1oSo0.48. The AUC provides a summarized scalar for topological characterization of brain networks independent of single threshold selection (Achard and Bullmore, 2007; He et al., 2009). The integrated AUC of network metric Y, which was computed over the threshold range of S1 to Sn with interval of ΔS, was expressed as follows: n1   Y AUC ¼ ∑ YðSk Þ þ YðSkþ1 Þ ΔS=2 k¼1

In the present study, S1 ¼ 0:1, Sn ¼ 0:48, and ΔS ¼ 0:01. The AUC metric has been used in previous studies and considered to be able to sensitively detect the topological differences between groups (Zhang et al., 2011). Subsequently, nonparametric permutation testing was applied to test between-group differences of the integrated AUC metrics using the above permutation framework. The differences of each graph characteristics with the p value lower than 0.05 were considered as significant.

Acknowledgments We thank Heather Shapiro of University of California at Davis for help with English editing. We are also grateful for the interest and participation of all the Chinese-chess grandmasters and masters, as well as the control volunteers. This work was supported by 973 Project (2012CB517901), the Natural Science Foundation of China (81301279, 61035006, 61125304, 81030027, and 81227002), the Specialized Research Fund for the Doctoral Program of Higher Education of China (20120185110028), and the Postdoctoral Science Foundation of China (2012M511824).

42

brain research 1558 (2014) 33–43

references

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. Achard, S., Bullmore, E., 2007. Efficiency and cost of economical brain functional networks. PLoS Comput. Biol. 3, e17. Albert, N.B., Robertson, E.M., Miall, R.C., 2009. The resting human brain and motor learning. Curr. Biol. 19, 1023–1027. Amidzic, O., Riehle, H.J., Fehr, T., Wienbruch, C., Elbert, T., 2001. Pattern of focal gamma-bursts in chess players. Nature 412, 603. Atherton, M., Zhuang, J., Bart, W.M., Hu, X., He, S., 2003. A functional MRI study of high-level cognition. I. The game of chess. Brain Res. Cogn. Brain Res. 16, 26–31. Bai, F., Shu, N., Yuan, Y., Shi, Y., Yu, H., Wu, D., Wang, J., Xia, M., He, Y., Zhang, Z., 2012. Topologically convergent and divergent structural connectivity patterns between patients with remitted geriatric depression and amnestic mild cognitive impairment. J. Neurosci. 32, 4307–4318. Bassett, D.S., Bullmore, E., 2006. Small-world brain networks. Neuroscientist 12, 512–523. Bilalic, M., Langner, R., Erb, M., Grodd, W., 2010. Mechanisms and neural basis of object and pattern recognition: a study with chess experts. J. Exp. Psychol. Gen. 139, 728–742. Bilalic, M., Langner, R., Ulrich, R., Grodd, W., 2011. Many faces of expertise: fusiform face area in chess experts and novices. J. Neurosci. 31, 10206–10214. Biswal, B., Yetkin, F.Z., Haughton, V.M., Hyde, J.S., 1995. Functional connectivity in the motor cortex of resting human brain using echo-planar MRI. Magn. Reson. Med. 34, 537–541. Bukach, C.M., Gauthier, I., Tarr, M.J., 2006. Beyond faces and modularity: the power of an expertise framework. Trends Cogn. Sci. 10, 159–166. Bullmore, E., Sporns, O., 2009. Complex brain networks: graph theoretical analysis of structural and functional systems. Nat. Rev. Neurosci. 10, 186–198. Campitelli, G., Gobet, F., Parker, A., 2005. Structure and stimulus familiarity: a study of memory in chess-players with functional magnetic resonance imaging. Span. J. Psychol. 8, 238–245. Campitelli, G., Gobet, F., Head, K., Buckley, M., Parker, A., 2007. Brain localization of memory chunks in chessplayers. Int. J. Neurosci. 117, 1641–1659. Charness, N., 1981. Aging and skilled problem solving. J. Exp. Psychol. Gen. 110, 21–38. Chase, W.G., Simon, H.A., 1973. Perception in chess. Cogn. Psychol. 4, 55–81. Connors, M.H., Burns, B.D., Campitelli, G., 2011. Expertise in complex decision making: the role of search in chess 70 years after de Groot. Cogn. Sci. 35, 1567–1579. De Groot, A.D., 1946. Het denken van den schaker. Noord Hollandsche, Amsterdam. de Groot, A.D., Gobet, F., 1996. Perception and Memory in Chess. Van Gorcum, Assen. Dickerson, K.C., Li, J., Delgado, M.R., 2011. Parallel contributions of distinct human memory systems during probabilistic learning. NeuroImage 55, 266–276. Duan, X., He, S., Liao, W., Liang, D., Qiu, L., Wei, L., Li, Y., Liu, C., Gong, Q., Chen, H., 2012a. Reduced caudate volume and enhanced striatal-DMN integration in chess experts. NeuroImage 60, 1280–1286. Duan, X., Liao, W., Liang, D., Qiu, L., Gao, Q., Liu, C., Gong, Q., Chen, H., 2012b. Large-scale brain networks in board game experts: insights from a domain-related task and task-free resting state. PLoS One 7, e32532.

Elo, A.E., 1978. The Rating of Chess Players, Past and Present. Arco, New York. Ericsson, K.A., Kintsch, W., 1995. Long-term working memory. Psychol. Rev. 102, 211–245. Fan, Y., Shi, F., Smith, J.K., Lin, W., Gilmore, J.H., Shen, D., 2011. Brain anatomical networks in early human brain development. Neuroimage 54, 1862–1871. Ferri, R., Rundo, F., Bruni, O., Terzano, M.G., Stam, C.J., 2007. Small-world network organization of functional connectivity of EEG slow-wave activity during sleep. Clin. Neurophysiol. 118, 449–456. Fox, M.D., Snyder, A.Z., Vincent, J.L., Raichle, M.E., 2007. Intrinsic fluctuations within cortical systems account for intertrial variability in human behavior. Neuron 56, 171–184. Gobet, F., Simon, H.A., 1998. Expert chess memory: revisiting the chunking hypothesis. Memory 6, 225–255. Gobet, F., Charness, N., 2006. Expertise in chess. In: Anders, E.K., Charness, N., Feltovich, P., Hoffman, R. (Eds.), The Cambridge Handbook of Expertise and Expert Performance. Cambridge University Press, Cambridge, pp. 523–538. Gong, G., He, Y., Concha, L., Lebel, C., Gross, D.W., Evans, A.C., Beaulieu, C., 2009. Mapping anatomical connectivity patterns of human cerebral cortex using in vivo diffusion tensor imaging tractography. Cereb. Cortex 19, 524–536. Graybiel, A.M., 2005. The basal ganglia: learning new tricks and loving it. Curr. Opin. Neurobiol. 15, 638–644. Greicius, M.D., Krasnow, B., Reiss, A.L., Menon, V., 2003. Functional connectivity in the resting brain: a network analysis of the default mode hypothesis. Proc. Natl. Acad. Sci. USA 100, 253–258. Haber, S.N., Calzavara, R., 2009. The cortico-basal ganglia integrative network: the role of the thalamus. Brain Res. Bull. 78, 69–74. Hagmann, P., Kurant, M., Gigandet, X., Thiran, P., Wedeen, V.J., Meuli, R., Thiran, J.P., 2007. Mapping human whole-brain structural networks with diffusion MRI. PLoS One 2, e597. Hampson, M., Driesen, N.R., Skudlarski, P., Gore, J.C., Constable, R. T., 2006a. Brain connectivity related to working memory performance. J. Neurosci. 26, 13338–13343. Hampson, M., Tokoglu, F., Sun, Z., Schafer, R.J., Skudlarski, P., Gore, J.C., Constable, R.T., 2006b. Connectivity-behavior analysis reveals that functional connectivity between left BA39 and Broca’s area varies with reading ability. NeuroImage 31, 513–519. 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., Dagher, A., Chen, Z., Charil, A., Zijdenbos, A., Worsley, K., Evans, A., 2009. Impaired small-world efficiency in structural cortical networks in multiple sclerosis associated with white matter lesion load. Brain 132, 3366–3379. He, Y., Evans, A., 2010. Graph theoretical modeling of brain connectivity. Curr. Opin. Neurol. 23, 341–350. Humphries, M.D., Gurney, K., Prescott, T.J., 2006. The brainstem reticular formation is a small-world, not scale-free, network. Proc. Biol. Sci. 273, 503–511. Lassiter, G.D., 2000. The relative contributions of recognition and search-evaluation processes to high-level chess performance: comment on Gobet and Simon. Psychol. Sci. 11, discussion 174 172–173. Lewis, C.M., Baldassarre, A., Committeri, G., Romani, G.L., Corbetta, M., 2009. Learning sculpts the spontaneous activity of the resting human brain. Proc. Natl. Acad. Sci. USA 106, 17558–17563. 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. Liao, W., Zhang, Z., Pan, Z., Mantini, D., Ding, J., Duan, X., Luo, C., Lu, G., Chen, H., 2010. Altered functional connectivity and

brain research 1558 (2014) 33–43

small-world in mesial temporal lobe epilepsy. PLoS One 5, e8525. Liao, W., Ding, J., Marinazzo, D., Xu, Q., Wang, Z., Yuan, C., Zhang, Z., Lu, G., Chen, H., 2011a. Small-world directed networks in the human brain: multivariate Granger causality analysis of restingstate fMRI. NeuroImage 54, 2683–2694. Liao, W., Zhang, Z., Pan, Z., Mantini, D., Ding, J., Duan, X., Luo, C., Wang, Z., Tan, Q., Lu, G., Chen, H., 2011b. Default mode network abnormalities in mesial temporal lobe epilepsy: a study combining fMRI and DTI. Hum. Brain Mapp. 32, 883–895. Liu, Y., Liang, M., Zhou, Y., He, Y., Hao, Y., Song, M., Yu, C., Liu, H., Liu, Z., Jiang, T., 2008. Disrupted small-world networks in schizophrenia. Brain 131, 945–961. Mennes, M., Zuo, X.N., Kelly, C., Di Martino, A., Zang, Y.F., Biswal, B., Castellanos, F.X., Milham, M.P., 2010. Linking inter-individual differences in neural activation and behavior to intrinsic brain dynamics. NeuroImage 54 (4), 2950–2959. Newman, M., 2003. The structure and function of complex networks. SIAM Rev. 45, 167–256. Nichelli, P., Grafman, J., Pietrini, P., Alway, D., Carton, J.C., Miletich, R., 1994. Brain activity in chess playing. Nature 369, 191. Onnela, J.P., Saramaki, J., Kertesz, J., Kaski, K., 2005. Intensity and coherence of motifs in weighted complex networks. Phys. Rev. E Stat. Nonlinear Soft Matter Phys. 71, 065103. Onofrj, M., Curatola, L., Valentini, G., Antonelli, M., Thomas, A., Fulgente, T., 1995. Non-dominant dorsal-prefrontal activation during chess problem solution evidenced by single photon emission computerized tomography (SPECT). Neurosci. Lett. 198, 169–172. Pasupathy, A., Miller, E.K., 2005. Different time courses of learning-related activity in the prefrontal cortex and striatum. Nature 433, 873–876. Poldrack, R.A., Clark, J., Pare-Blagoev, E.J., Shohamy, D., Creso Moyano, J., Myers, C., Gluck, M.A., 2001. Interactive memory systems in the human brain. Nature 414, 546–550. Poldrack, R.A., Rodriguez, P., 2004. How do memory systems interact? Evidence from human classification learning. Neurobiol. Learn. Mem. 82, 324–332. Ponten, S.C., Bartolomei, F., Stam, C.J., 2007. Small-world networks and epilepsy: graph theoretical analysis of intracerebrally recorded mesial temporal lobe seizures. Clin. Neurophysiol. 118, 918–927. Raichle, M.E., 2010. The brain’s dark energy. Sci. Am. 302, 44–49. Reynolds, R.I., 1982. Search heuristics of chess players of different calibers. Am. J. Psychol. 95, 383–392. Rubinov, M., Sporns, O., 2010. Complex network measures of brain connectivity: uses and interpretations. NeuroImage 52, 1059–1069. Salvador, R., Suckling, J., Coleman, M.R., Pickard, J.D., Menon, D., Bullmore, E., 2005. Neurophysiological architecture of functional magnetic resonance images of human brain. Cereb. Cortex 15, 1332–1342. Sanz-Arigita, E.J., Schoonheim, M.M., Damoiseaux, J.S., Rombouts, S.A., Maris, E., Barkhof, F., Scheltens, P., Stam, C.J., 2010. Loss of ‘small-world’ networks in Alzheimer’s disease: graph analysis of FMRI resting-state functional connectivity. PLoS One 5, e13788. Smit, D.J., Stam, C.J., Posthuma, D., Boomsma, D.I., de Geus, E.J., 2008. Heritability of “small-world” networks in the brain: a graph theoretical analysis of resting-state EEG functional connectivity. Hum. Brain Mapp. 29, 1368–1378. Sporns, O., Chialvo, D.R., Kaiser, M., Hilgetag, C.C., 2004. Organization, development and function of complex brain networks. Trends Cogn. Sci. 8, 418–425. Stam, C.J., Jones, B.F., Nolte, G., Breakspear, M., Scheltens, P., 2007. Small-world networks and functional connectivity in Alzheimer’s disease. Cereb. Cortex 17, 92–99. Stam, C.J., de Haan, W., Daffertshofer, A., Jones, B.F., Manshanden, I., van Cappellen van Walsum, A.M., Montez, T.,

43

Verbunt, P., de Munck, J.C., van Dijk, B.W., Berendse, H.W., Scheltens, P., 2009. Graph theoretical analysis of magnetoencephalographic functional connectivity in Alzheimer’s disease. Brain 132, 213–224. Supekar, K., Musen, M., Menon, V., 2009. Development of largescale functional brain networks in children. PLoS Biol. 7, e1000157. 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., Stam, C.J., Boersma, M., Hulshoff Pol, H.E., 2008. Small-world and scale-free organization of voxel-based resting-state functional connectivity in the human brain. NeuroImage 43, 528–539. 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 Maas, H.L., Wagenmakers, E.J., 2005. A psychometric analysis of chess expertise. Am. J. Psychol. 118, 29–60. Voermans, N.C., Petersson, K.M., Daudey, L., Weber, B., Van Spaendonck, K.P., Kremer, H.P., Fernandez, G., 2004. Interaction between the human hippocampus and the caudate nucleus during route recognition. Neuron 43, 427–435. Voss, M.W., Prakash, R.S., Erickson, K.I., Basak, C., Chaddock, L., Kim, J.S., Alves, H., Heo, S., Szabo, A.N., White, S.M., Wojcicki, T.R., Mailey, E.L., Gothe, N., Olson, E.A., McAuley, E., Kramer, A.F., 2010. Plasticity of brain networks in a randomized intervention trial of exercise training in older adults. Front. Aging Neurosci. 2, p. 32. Wan, X., Nakatani, H., Ueno, K., Asamizuya, T., Cheng, K., Tanaka, K., 2011. The neural basis of intuitive best next-move generation in board game experts. Science 331, 341–346. Wang, J., Wang, L., Zang, Y., Yang, H., Tang, H., Gong, Q., Chen, Z., Zhu, C., He, Y., 2009a. Parcellation-dependent small-world brain functional networks: a resting-state fMRI study. Hum. Brain Mapp. 30, 1511–1523. Wang, L., Zhu, C., He, Y., Zang, Y., Cao, Q., Zhang, H., Zhong, Q., Wang, Y., 2009b. Altered small-world brain functional networks in children with attention-deficit/hyperactivity disorder. Hum. Brain Mapp. 30, 638–649. Wang, Q., Su, T.P., Zhou, Y., Chou, K.H., Chen, I.Y., Jiang, T., Lin, C.P., 2012. Anatomical insights into disrupted small-world networks in schizophrenia. NeuroImage 59, 1085–1093. Wang, Y., Tran, H.D., Liao, L., Duan, X., Kaner, R.B., 2010. Nanoscale morphology, dimensional control, and electrical properties of oligoanilines. J. Am. Chem. Soc. 132, 10365–10373. Watts, D.J., Strogatz, S.H., 1998. Collective dynamics of ‘smallworld’ networks. Nature 393, 440–442. Yin, H.H., Ostlund, S.B., Balleine, B.W., 2008. Reward-guided learning beyond dopamine in the nucleus accumbens: the integrative functions of cortico-basal ganglia networks. Eur. J. Neurosci. 28, 1437–1448. Yu, Q., Sui, J., Rachakonda, S., He, H., Gruner, W., Pearlson, G., Kiehl, K.A., Calhoun, V.D., 2011. Altered topological properties of functional network connectivity in schizophrenia during resting state: a small-world brain network study. PLoS One 6, e25423. Zalesky, A., Fornito, A., Bullmore, E.T., 2010. Network-based statistic: identifying differences in brain networks. NeuroImage 53, 1197–1207. Zhang, D., Raichle, M.E., 2010. Disease and the brain’s dark energy. Nat. Rev. Neurol. 6, 15–28. Zhang, J., Wang, J., Wu, Q., Kuang, W., Huang, X., He, Y., Gong, Q., 2011. Disrupted brain connectivity networks in drug-naive, firstepisode major depressive disorder. Biol. Psychiatry 70, 334–342.