Brain Imaging and Behavior DOI 10.1007/s11682-013-9279-3
ORIGINAL RESEARCH
Global and local brain network reorganization in attention-deficit/hyperactivity disorder Pan Lin & Jubao Sun & Gang Yu & Ying Wu & Yong Yang & Meilin Liang & Xin Liu
# Springer Science+Business Media New York 2013
Abstract Brain is a complex network with an anatomical and functional organization. The differences in brain organization of those with attention-deficit/hyperactivity disorder (ADHD) are still not well understood. Here, we study brain organization in ADHD subjects using a complex network derived from resting-state functional magnetic resonance imaging (fMRI) data of ADHD and normal subjects. Our results reveal that the brain networks of ADHD subjects are reorganized compared to those without ADHD in global and local brain functional networks. We find that the ADHD subjects show decreasing brain network integration and increasing brain network segregation. More interestingly, we find similarities of brain topology properties between local and global brain networks. Our finding indicates that cognitive dysfunction in ADHD is probably associated with disrupted global and local brain Pan Lin and Jubao Sun equally contributing authors P. Lin (*) : M. Liang : X. Liu Key Laboratory of Biomedical Information Engineering of Education Ministry, Institute of Biomedical Engineering, Xi’an Jiaotong University, Xi’an 710049, China e-mail: [email protected] J. Sun MRI Center, The First Affiliated Hospital of Henan University of Science & Technology, Luoyang, Henan 471000, China G. Yu School of Info-Physics and Geomatics Engineering, Central South University, Changsha 410083, China Y. Wu State Key Laboratory for Strength and Vibration of Mechanical Structures, School of Aerospace, Xi’an Jiaotong University, Xi’an 710049, China Y. Yang School of Information Technology, Jiangxi University of Finance and Economics, Nanchang 330013, China
network topological properties. Our results can help us understand the pathophysiological mechanism of ADHD and serve as a sensitive and specific biomarker of ADHD. Keyword Attention-deficit/hyperactivity disorder (ADHD) . Resting-state functional connectivity . Functional magnetic resonance imaging . Complex networks
Introduction Attention-deficit/hyperactivity disorder (ADHD) is one of the most commonly diagnosed childhood neuropsychiatry disorders and is characterized by inattention, hyperactivity and impulsivity with symptoms sometimes starting before 7 years of age. ADHD is affecting 3~5 % of children globally, and 30~50 % of those children will still have symptoms in their adult life (Balint et al. 2008). Children with ADHD are found easily distracted, have difficultly focusing on one task, need to constantly be in motion, and are impatient, which all affect their academic performance and social life. Neuroimaging is very useful for brain cognition and mental disorder studies (Diwadkar et al. 2008; Parisi et al. 2012; Stevens et al. 2012; Zhang et al. 2012; Peng et al. 2013; Leavitt et al. 2012; Robinson et al. 2009). Neuroimaging studies demonstrate that the ADHD dysfunction is associated with abnormalities in frontal, parietal, temporal, occipital, and sub-cortical regions of the brain (Seidman et al. 2004; Bush et al. 2005; Seidman et al. 2005). These abnormal brain regions can be used to construct various functional networks, such as fronto-striatal (Castellanos et al. 2008), fronto-parietal (Dickstein et al. 2006) and fronto-temporal-parietal circuits (Smith et al. 2006). The disruptions of functional subsystems constructed by these abnormal brain regions may lead to the
Brain Imaging and Behavior
functional network deficit in ADHD. (Seidman et al. 2004; Bush et al. 2005). The brain is a complex dynamic system (Bullmore and Sporns 2009). Recent developments in neuroimaging have revealed the complex brain organization of large-scale networks. A large-scale brain functional network shapes brain function and facilitates cognitive processing. Large-scale functional brain networks can be identified using electroencephalography (EEG), magnetoencephalography (MEG), and functional magnetic resonance imaging (fMRI) (Bassett et al. 2009; Bullmore and Sporns 2009; Sporns 2011). This complex network provides powerful tools to examine how functional connectivity and information integration relate to human behavior, and how network organization is altered in neurodegenerative diseases (Bullmore and Sporns 2009). The complex network approach is increasingly used to investigate the topological properties of both structural and functional networks in the human brain. (Salvador et al. 2005; Achard et al. 2006; Bassett and Bullmore 2006; Ferri et al. 2007; Micheloyannis et al. 2009). Numerous studies have demonstrated that functional brain complex network properties could be affected by brain diseases such as schizophrenia (Zhang et al. 2012), Alzheimer’s disease (Stam et al. 2007) and ADHD disease (Wang et al. 2009; Konrad and Eickhoff 2010). Understanding brain network organization could facilitate better diagnosis of brain disorders and guide treatment for psychiatric disorders. A variety of complex network methods have been used to investigate the entire brain network for diagnosing brain disorders (Stam et al. 2007; Meunier et al. 2009; Micheloyannis et al. 2009; Stam 2010; Xia and He 2011; Castellanos and Proal 2012; Hwang et al. 2013). However, the brain network is a complex dynamic network in which information is continuously processed and transported between spatially distributed but functionally linked sub-networks or individual brain regions (Andrews-Hanna et al. 2010; Power et al. 2011). Accumulating evidence shows that sub-networks of intrinsic resting state brain influence cognitive behavior (van den Heuvel and Hulshoff Pol 2010; Castellanos and Proal 2012; Vitevitch et al. 2012). Recent studies based on resting state fMRI data demonstrate that the functional connectivity is abnormal in the brain network of ADHD patients (Wang et al. 2009; Fair et al. 2010; Castellanos and Proal 2012). For example, the ADHD group shows decreasing functional connectivity compared to the healthy control group in fronto-striatal, fronto-parietal and fronto-cerebellar sub-networks (Cao et al. 2006; Castellanos et al. 2008; Cubillo et al. 2010). Furthermore, Resting state functional connectivity studies indicate that ADHD group brain dysfunction involves large-scale brain sub-networks such as frontoparietal network, attention network, motor network, and the default network (Castellanos and Proal 2012). Most of the studies on sub-networks refer to the default mode
network (DMN) of the brain (Raichle et al. 2001; Lin et al. 2011; De Pisapia et al. 2012). Several resting state functional connectivity studies suggest that ADHD is associated with a DMN disorder (Greicius 2008). So, it is more important to characterize the relationship between the local brain network topology properties and brain function for understanding the ADHD neuropsychological mechanism. It is still unclear how the global or local brain network deficits link to dysfunction in ADHD. Is the local brain network reorganization associated with ADHD disorder? To better understand whether global and local topology properties shift in ADHD, we used the complex network to investigate the difference of the resting state brain network topologies between ADHD subjects and normal control subjects. We hypothesized that global and local functional brain organization would be altered in ADHD. Further, we sought to find out whether global and local brain network topology properties obtained from resting state fMRI data might provide a sensitive and specific biomarker in ADHD.
Materials and methods Subjects and MRI acquisition fMRI data are extracted from the open-access ‘1000 Functional Connectomes Project’ (http://fcon_1000.projects. nitrc.org/) in which resting-state fMRI scans have been released by M.P. Milham and F.X. Castellanos in December, 2009. These data are acquiesced at resting state by a 3T Siemens scanner. There are 25 ADHD subjects in group A (19M/4F,ages 20–50), and 84 normal subjects in group B (43M/41F,ages 7–49). The patients recruited in this dataset were evaluated with the clinical interview DSM-IV (SCID), Checklist-90-Revised (SCL-90-R) and Adult ADHD Clinical Diagnostic Scale (ACDS). Image scans contain 39 slices and 192 time points; TR is 2 s, TE = 25 ms, flip angle = 90, matrix = 64×64; FOV = 192 mm; voxel size = 3×3×3 mm and the last time is 390 s. We selected 24 ADHD subjects from group A, 19 male and 5 female, and the mean age is 34.87. We also selected 24 normal subjects from group B, 18 male and 6 female, and the mean age is 34.65. Using two-sample t-tests, the result shows no significant difference in age and gender (p >0.05, see Table 1). Table 1 Demographics of the subjects
Gender(male) Age(years)
Control
ADHD
P-value
18 34.65±9.15
19 34.87±9.77
0.5348 0.7379
• The P-value was obtained by two-sample two-tailed t -test
Brain Imaging and Behavior
FMRI data preprocessing
Table 2 Cortical and sub-cortical regions defined in automated anatomical labeling template image in standard stereotaxic space
The functional images are preprocessed based on AFNI (http://afni.nimh.nih.gov/afni/) (Cox 1996) and FSL’s software Library (http://www.fmrib.ox.ac.uk/fsl/). The first four volumes are excluded from analysis for initial stabilization of the fMRI signal. For each subject, motion correction is performed through a 3D image realignment with the AFNI program 3dvolreg function, which uses a weighted least squares rigid-body registration algorithm. Echo planar imaging (EPI) images were motion and slicetime corrected, and spatially smoothed using a Gaussian kernel of 6 mm Full width at half maximum (FWHM). The temporal band-pass filtering (0.005 < f < 0.1 Hz) is performed in order to reduce the effects of low-frequency drift and highfrequency physiological noise. After eliminating redundant information of Cerebral Spinal Fluid (CSF) and white matter, fMRI data are further spatially normalized to the Montreal Neurological Institute (MNI) EPI template and resampled to a 3 mm cubic voxel.
Region
Abbreviation
Amygdala left/right Lingual gyrus left/right Posterior cingulum gyrus left/right Rolandic operculum left/right
AMY.L/AMY.R LIN.L/LIN.R PCG.L/PCG.R ROP.L/ROP.R
Angular gyrus left/right Cuneus left/right Inferior occipital gyrus left/right Supplementary motor area left/right Calcarine cortex left/right Olfactory left/right Cerebelum_3 left/right Orbitofrontal gyrus (medial) left/right Pallidum left/right Paracentral lobule left/right Orbitofrontal gyrus (middle) left/right ParaHippocampal gyrus left/right Cerebelum_7b left/right left/right Inferior parietal lobule left/right Superior frontal gyrus(medial) left/right Superior parietal gyrus left/right Cerebelum_9 left/right Cerebelum_Crus1 left/right
ANG.L/ANG.R CUN.L/CUN.R IOC.L/IOC.R SuMo.L/SuMo.R CAL.L/CAL.R ALF.L/ALF.R Cer3.L/Cer3.R OFMe.L/OFMe.R PAL.L/PAL.R PARCL.L/PARCL.R OFMi.L/OFMi.R PAHI.L/PAHI.R Ce7b.L/Ce7b.R IPL.L/IPL.R SFMe.L/SFMe.R SPG.L/SPG.R Cer9.L/Cer9.R CerC1.L/CerC1.R
Precentral gyrus left/right Heschl gyrus left/right Vermis_1_2* Vermis_4_5* Vermis_10* Vermis_8* Anterior cingulate gyrus (L*) Putamen (L*) Inferior frontal gyrus (opercular) left/right Middle occipital gyrus left/right SupraMarginal gyrus left/right Caudate left/right Orbitofrontal cortex (inferior) left/right Superior occipital gyrus left/right Inferior temporal gyrus left/right Cerebelum_10 left/right Inferior frontal gyrus(triangular) left/right Middle temporal gyrus left/right Temporal pole (middle) left/right
PrG.L/PrG.R HES.L/HES.R Ver12 Ver45 Ver10 Ver8 ACG PUT IFOp.L/IFOp.R MOC.L/MOC.R SuMa.L/SuMa.R CAU.L/CAU.R OFI.L/OFI.R SuOc.L/SuOc.R INT.L/INT.R Ce10.L/Ce10.R IFT.L/IFT.R MiT.L/MiT.R TPMi.L/TPMi.R
Cerebelum_4_5 left/right Middle frontal gyrus left/right Temporal pole (superior) left/right Cerebelum_6 left/right Superior temporal gyrus left/right Superior frontal gyrus left/right Thalamus left/right
Ce45.L/Ce45.R MiF.L/MiF.R TPS.L/TPS.R Cer6.L/Cer6.R STG.L/STG.R SFG.L/SFG.R THA.L/THA.R
Construction of brain functional connectivity correlation matrix In the present study, we construct brain functional networks using the automated anatomical labeling (AAL) atlas (Tzourio-Mazoyer et al. 2002) to parcellate the brain into 108 regions of interest (ROIs). The names of the ROIs and their corresponding abbreviations are listed in Table 2. For each participant’s resting state fMRI data, ROIs mean time series are calculated by taking the mean of the voxel time series within each region. The Pearson’s correlation coefficients are computed between each pair of ROIs for each subject, then the 108×108 correlation matrix for each subject is obtained. Correlation coefficients represent the functional connectivity strength between the ROIs. For further statistical analysis, a Fisher’s r-to-z transformation is applied to improve the normality of the correlation coefficients. Complex network analysis Graph theory is the natural framework for the exact mathematical representation of complex networks. Formally, a complex network can be represented as a graph by G(N, K), with N denoting the number of nodes and K denoting the number of edges in graph G. (Rubinov and Sporns 2010). To date, most brain network studies have investigated topology properties using binarized graphs. The related complex network topology measures include degree, global efficiency and local efficiency, modularity, clustering coefficient, and the shortest Path (see Appendix). First, we investigated the topological properties of brain functional networks as a function of threshold or cost. It is
Brain Imaging and Behavior Table 2 (continued)
choose the cost from 0.1 to 0.4 using increments of 0.05 and then estimate the network properties at each cost value. The binarized matrices are obtained through thresholding and the further analysis is performed based on the binarized matrices for each subject. The networks have been analyzed based on the Matlab BCT toolbox and our own program (Rubinov and Sporns 2010), (http://www.brain-connectivity-toolbox.net).
properties between the ADHD and control groups. The results are shown in Fig. 1. The results reveal that there are significant differences in global efficiency between ADHD and control groups over the range of cost 0.1~0.3 (p
ORIGINAL RESEARCH
Global and local brain network reorganization in attention-deficit/hyperactivity disorder Pan Lin & Jubao Sun & Gang Yu & Ying Wu & Yong Yang & Meilin Liang & Xin Liu
# Springer Science+Business Media New York 2013
Abstract Brain is a complex network with an anatomical and functional organization. The differences in brain organization of those with attention-deficit/hyperactivity disorder (ADHD) are still not well understood. Here, we study brain organization in ADHD subjects using a complex network derived from resting-state functional magnetic resonance imaging (fMRI) data of ADHD and normal subjects. Our results reveal that the brain networks of ADHD subjects are reorganized compared to those without ADHD in global and local brain functional networks. We find that the ADHD subjects show decreasing brain network integration and increasing brain network segregation. More interestingly, we find similarities of brain topology properties between local and global brain networks. Our finding indicates that cognitive dysfunction in ADHD is probably associated with disrupted global and local brain Pan Lin and Jubao Sun equally contributing authors P. Lin (*) : M. Liang : X. Liu Key Laboratory of Biomedical Information Engineering of Education Ministry, Institute of Biomedical Engineering, Xi’an Jiaotong University, Xi’an 710049, China e-mail: [email protected] J. Sun MRI Center, The First Affiliated Hospital of Henan University of Science & Technology, Luoyang, Henan 471000, China G. Yu School of Info-Physics and Geomatics Engineering, Central South University, Changsha 410083, China Y. Wu State Key Laboratory for Strength and Vibration of Mechanical Structures, School of Aerospace, Xi’an Jiaotong University, Xi’an 710049, China Y. Yang School of Information Technology, Jiangxi University of Finance and Economics, Nanchang 330013, China
network topological properties. Our results can help us understand the pathophysiological mechanism of ADHD and serve as a sensitive and specific biomarker of ADHD. Keyword Attention-deficit/hyperactivity disorder (ADHD) . Resting-state functional connectivity . Functional magnetic resonance imaging . Complex networks
Introduction Attention-deficit/hyperactivity disorder (ADHD) is one of the most commonly diagnosed childhood neuropsychiatry disorders and is characterized by inattention, hyperactivity and impulsivity with symptoms sometimes starting before 7 years of age. ADHD is affecting 3~5 % of children globally, and 30~50 % of those children will still have symptoms in their adult life (Balint et al. 2008). Children with ADHD are found easily distracted, have difficultly focusing on one task, need to constantly be in motion, and are impatient, which all affect their academic performance and social life. Neuroimaging is very useful for brain cognition and mental disorder studies (Diwadkar et al. 2008; Parisi et al. 2012; Stevens et al. 2012; Zhang et al. 2012; Peng et al. 2013; Leavitt et al. 2012; Robinson et al. 2009). Neuroimaging studies demonstrate that the ADHD dysfunction is associated with abnormalities in frontal, parietal, temporal, occipital, and sub-cortical regions of the brain (Seidman et al. 2004; Bush et al. 2005; Seidman et al. 2005). These abnormal brain regions can be used to construct various functional networks, such as fronto-striatal (Castellanos et al. 2008), fronto-parietal (Dickstein et al. 2006) and fronto-temporal-parietal circuits (Smith et al. 2006). The disruptions of functional subsystems constructed by these abnormal brain regions may lead to the
Brain Imaging and Behavior
functional network deficit in ADHD. (Seidman et al. 2004; Bush et al. 2005). The brain is a complex dynamic system (Bullmore and Sporns 2009). Recent developments in neuroimaging have revealed the complex brain organization of large-scale networks. A large-scale brain functional network shapes brain function and facilitates cognitive processing. Large-scale functional brain networks can be identified using electroencephalography (EEG), magnetoencephalography (MEG), and functional magnetic resonance imaging (fMRI) (Bassett et al. 2009; Bullmore and Sporns 2009; Sporns 2011). This complex network provides powerful tools to examine how functional connectivity and information integration relate to human behavior, and how network organization is altered in neurodegenerative diseases (Bullmore and Sporns 2009). The complex network approach is increasingly used to investigate the topological properties of both structural and functional networks in the human brain. (Salvador et al. 2005; Achard et al. 2006; Bassett and Bullmore 2006; Ferri et al. 2007; Micheloyannis et al. 2009). Numerous studies have demonstrated that functional brain complex network properties could be affected by brain diseases such as schizophrenia (Zhang et al. 2012), Alzheimer’s disease (Stam et al. 2007) and ADHD disease (Wang et al. 2009; Konrad and Eickhoff 2010). Understanding brain network organization could facilitate better diagnosis of brain disorders and guide treatment for psychiatric disorders. A variety of complex network methods have been used to investigate the entire brain network for diagnosing brain disorders (Stam et al. 2007; Meunier et al. 2009; Micheloyannis et al. 2009; Stam 2010; Xia and He 2011; Castellanos and Proal 2012; Hwang et al. 2013). However, the brain network is a complex dynamic network in which information is continuously processed and transported between spatially distributed but functionally linked sub-networks or individual brain regions (Andrews-Hanna et al. 2010; Power et al. 2011). Accumulating evidence shows that sub-networks of intrinsic resting state brain influence cognitive behavior (van den Heuvel and Hulshoff Pol 2010; Castellanos and Proal 2012; Vitevitch et al. 2012). Recent studies based on resting state fMRI data demonstrate that the functional connectivity is abnormal in the brain network of ADHD patients (Wang et al. 2009; Fair et al. 2010; Castellanos and Proal 2012). For example, the ADHD group shows decreasing functional connectivity compared to the healthy control group in fronto-striatal, fronto-parietal and fronto-cerebellar sub-networks (Cao et al. 2006; Castellanos et al. 2008; Cubillo et al. 2010). Furthermore, Resting state functional connectivity studies indicate that ADHD group brain dysfunction involves large-scale brain sub-networks such as frontoparietal network, attention network, motor network, and the default network (Castellanos and Proal 2012). Most of the studies on sub-networks refer to the default mode
network (DMN) of the brain (Raichle et al. 2001; Lin et al. 2011; De Pisapia et al. 2012). Several resting state functional connectivity studies suggest that ADHD is associated with a DMN disorder (Greicius 2008). So, it is more important to characterize the relationship between the local brain network topology properties and brain function for understanding the ADHD neuropsychological mechanism. It is still unclear how the global or local brain network deficits link to dysfunction in ADHD. Is the local brain network reorganization associated with ADHD disorder? To better understand whether global and local topology properties shift in ADHD, we used the complex network to investigate the difference of the resting state brain network topologies between ADHD subjects and normal control subjects. We hypothesized that global and local functional brain organization would be altered in ADHD. Further, we sought to find out whether global and local brain network topology properties obtained from resting state fMRI data might provide a sensitive and specific biomarker in ADHD.
Materials and methods Subjects and MRI acquisition fMRI data are extracted from the open-access ‘1000 Functional Connectomes Project’ (http://fcon_1000.projects. nitrc.org/) in which resting-state fMRI scans have been released by M.P. Milham and F.X. Castellanos in December, 2009. These data are acquiesced at resting state by a 3T Siemens scanner. There are 25 ADHD subjects in group A (19M/4F,ages 20–50), and 84 normal subjects in group B (43M/41F,ages 7–49). The patients recruited in this dataset were evaluated with the clinical interview DSM-IV (SCID), Checklist-90-Revised (SCL-90-R) and Adult ADHD Clinical Diagnostic Scale (ACDS). Image scans contain 39 slices and 192 time points; TR is 2 s, TE = 25 ms, flip angle = 90, matrix = 64×64; FOV = 192 mm; voxel size = 3×3×3 mm and the last time is 390 s. We selected 24 ADHD subjects from group A, 19 male and 5 female, and the mean age is 34.87. We also selected 24 normal subjects from group B, 18 male and 6 female, and the mean age is 34.65. Using two-sample t-tests, the result shows no significant difference in age and gender (p >0.05, see Table 1). Table 1 Demographics of the subjects
Gender(male) Age(years)
Control
ADHD
P-value
18 34.65±9.15
19 34.87±9.77
0.5348 0.7379
• The P-value was obtained by two-sample two-tailed t -test
Brain Imaging and Behavior
FMRI data preprocessing
Table 2 Cortical and sub-cortical regions defined in automated anatomical labeling template image in standard stereotaxic space
The functional images are preprocessed based on AFNI (http://afni.nimh.nih.gov/afni/) (Cox 1996) and FSL’s software Library (http://www.fmrib.ox.ac.uk/fsl/). The first four volumes are excluded from analysis for initial stabilization of the fMRI signal. For each subject, motion correction is performed through a 3D image realignment with the AFNI program 3dvolreg function, which uses a weighted least squares rigid-body registration algorithm. Echo planar imaging (EPI) images were motion and slicetime corrected, and spatially smoothed using a Gaussian kernel of 6 mm Full width at half maximum (FWHM). The temporal band-pass filtering (0.005 < f < 0.1 Hz) is performed in order to reduce the effects of low-frequency drift and highfrequency physiological noise. After eliminating redundant information of Cerebral Spinal Fluid (CSF) and white matter, fMRI data are further spatially normalized to the Montreal Neurological Institute (MNI) EPI template and resampled to a 3 mm cubic voxel.
Region
Abbreviation
Amygdala left/right Lingual gyrus left/right Posterior cingulum gyrus left/right Rolandic operculum left/right
AMY.L/AMY.R LIN.L/LIN.R PCG.L/PCG.R ROP.L/ROP.R
Angular gyrus left/right Cuneus left/right Inferior occipital gyrus left/right Supplementary motor area left/right Calcarine cortex left/right Olfactory left/right Cerebelum_3 left/right Orbitofrontal gyrus (medial) left/right Pallidum left/right Paracentral lobule left/right Orbitofrontal gyrus (middle) left/right ParaHippocampal gyrus left/right Cerebelum_7b left/right left/right Inferior parietal lobule left/right Superior frontal gyrus(medial) left/right Superior parietal gyrus left/right Cerebelum_9 left/right Cerebelum_Crus1 left/right
ANG.L/ANG.R CUN.L/CUN.R IOC.L/IOC.R SuMo.L/SuMo.R CAL.L/CAL.R ALF.L/ALF.R Cer3.L/Cer3.R OFMe.L/OFMe.R PAL.L/PAL.R PARCL.L/PARCL.R OFMi.L/OFMi.R PAHI.L/PAHI.R Ce7b.L/Ce7b.R IPL.L/IPL.R SFMe.L/SFMe.R SPG.L/SPG.R Cer9.L/Cer9.R CerC1.L/CerC1.R
Precentral gyrus left/right Heschl gyrus left/right Vermis_1_2* Vermis_4_5* Vermis_10* Vermis_8* Anterior cingulate gyrus (L*) Putamen (L*) Inferior frontal gyrus (opercular) left/right Middle occipital gyrus left/right SupraMarginal gyrus left/right Caudate left/right Orbitofrontal cortex (inferior) left/right Superior occipital gyrus left/right Inferior temporal gyrus left/right Cerebelum_10 left/right Inferior frontal gyrus(triangular) left/right Middle temporal gyrus left/right Temporal pole (middle) left/right
PrG.L/PrG.R HES.L/HES.R Ver12 Ver45 Ver10 Ver8 ACG PUT IFOp.L/IFOp.R MOC.L/MOC.R SuMa.L/SuMa.R CAU.L/CAU.R OFI.L/OFI.R SuOc.L/SuOc.R INT.L/INT.R Ce10.L/Ce10.R IFT.L/IFT.R MiT.L/MiT.R TPMi.L/TPMi.R
Cerebelum_4_5 left/right Middle frontal gyrus left/right Temporal pole (superior) left/right Cerebelum_6 left/right Superior temporal gyrus left/right Superior frontal gyrus left/right Thalamus left/right
Ce45.L/Ce45.R MiF.L/MiF.R TPS.L/TPS.R Cer6.L/Cer6.R STG.L/STG.R SFG.L/SFG.R THA.L/THA.R
Construction of brain functional connectivity correlation matrix In the present study, we construct brain functional networks using the automated anatomical labeling (AAL) atlas (Tzourio-Mazoyer et al. 2002) to parcellate the brain into 108 regions of interest (ROIs). The names of the ROIs and their corresponding abbreviations are listed in Table 2. For each participant’s resting state fMRI data, ROIs mean time series are calculated by taking the mean of the voxel time series within each region. The Pearson’s correlation coefficients are computed between each pair of ROIs for each subject, then the 108×108 correlation matrix for each subject is obtained. Correlation coefficients represent the functional connectivity strength between the ROIs. For further statistical analysis, a Fisher’s r-to-z transformation is applied to improve the normality of the correlation coefficients. Complex network analysis Graph theory is the natural framework for the exact mathematical representation of complex networks. Formally, a complex network can be represented as a graph by G(N, K), with N denoting the number of nodes and K denoting the number of edges in graph G. (Rubinov and Sporns 2010). To date, most brain network studies have investigated topology properties using binarized graphs. The related complex network topology measures include degree, global efficiency and local efficiency, modularity, clustering coefficient, and the shortest Path (see Appendix). First, we investigated the topological properties of brain functional networks as a function of threshold or cost. It is
Brain Imaging and Behavior Table 2 (continued)
choose the cost from 0.1 to 0.4 using increments of 0.05 and then estimate the network properties at each cost value. The binarized matrices are obtained through thresholding and the further analysis is performed based on the binarized matrices for each subject. The networks have been analyzed based on the Matlab BCT toolbox and our own program (Rubinov and Sporns 2010), (http://www.brain-connectivity-toolbox.net).
properties between the ADHD and control groups. The results are shown in Fig. 1. The results reveal that there are significant differences in global efficiency between ADHD and control groups over the range of cost 0.1~0.3 (p
