Alterations in regional homogeneity of baseline brain activity in pediatric temporal lobe epilepsy

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –2 29

available at www.sciencedirect.com

www.elsevier.com/locate/brainres

Research Report

Alterations in regional homogeneity of baseline brain activity in pediatric temporal lobe epilepsy Katariina Mankinen a,⁎, Xiang-Yu Long b , Jyri-Johan Paakki c , Marika Harila d , Seppo Rytky e , Osmo Tervonen c , Juha Nikkinen c , Tuomo Starck c , Jukka Remes c , Heikki Rantala a , Yu-Feng Zang b , Vesa Kiviniemi c a

Department of Pediatrics, Oulu University Hospital, Oulu, Finland State Key Laboratory of Cognitive Neuroscience, Beijing Normal University, China c Clinic of Diagnostic Radiology, Oulu University Hospital, Oulu, Finland d Department of Neurology, Oulu University Hospital, Finland e Department of Clinical Neurophysiology, Oulu University Hospital, Finland b

A R T I C LE I N FO

AB S T R A C T

Article history:

Recent findings on intracortical EEG measurements show that the synchrony of localized

Accepted 2 December 2010

neuronal networks is altered in epileptogenesis, leading to generalized seizure activity via

Available online 10 December 2010

connector hubs in the neuronal networks. Regional homogeneity (ReHo) analysis of blood oxygen level-dependent (BOLD) signals has demonstrated localized signal synchrony and

Keywords:

disease-related alterations in a number of instances. We wanted to find out whether the

Regional homogeneity

ReHo of resting-state activity can be used to detect regional signal synchrony alterations in

Spontaneous state activity

children with non-lesional temporal lobe epilepsy (TLE). Twenty-one TLE patients were

Temporal lobe epilepsy

compared with age and gender-matched healthy controls. Significantly increased ReHo was

Interictal electroencephalogram

discovered in the posterior cingulate gyrus and the right medial temporal lobe of the

Resting-state functional magnetic

patients, and they also had significantly decreased ReHo in the cerebellum compared with

resonance imaging

the healthy controls. However, the alterations in ReHo differed between the patients with normal and abnormal interictal EEGs, the latter showing significantly increased ReHo in the right fusiform gyrus and significantly decreased ReHo in the right medial frontal gyrus relative to the controls, while those with normal EEGs had significantly increased ReHo in the right inferior temporal gyrus and the left posterior cingulate gyrus. We conclude that altered BOLD signal synchrony can be detected in the cerebral and cerebellar cortices of children with TLE even in the absence of interictal EEG abnormalities. © 2010 Elsevier B.V. All rights reserved.

⁎ Corresponding author. Oulu University Hospital, Department of Pediatrics, PB 29, 90014 Oulu, Finland. Fax: +358 8 3155559. E-mail address: [email protected] (K. Mankinen). Abbreviations: ReHo, regional homogeneity; BOLD, blood oxygen level-dependent; TLE, temporal lobe epilepsy; fMRI, functional magnetic resonance imaging; EEG, electroencephalogram; ADHD, attention deficit hyperactivity disorder; IQ, intelligence quotient; KCC, Kendall's coefficient of concordance; PET, positron emission tomography; AED, antiepileptic drug 0006-8993/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.brainres.2010.12.004

222

1.

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –22 9

Introduction

It has been suggested that interictal spike activity results from the summated activity of hypersynchronous epileptic tissue (Matsumoto and Ajmone Marsan, 1964), and that ictal epileptic activity may be preceded by preictal de- or hypersynchronization (Lopes da Silva et al., 2003; Wendling et al., 2005). Strong synchronization foci seem to be particularly involved in the generation of interictal activity in the temporal lobe (Ortega et al., 2008), and mesial temporal lobe seizures have been characterized by increased local coherence as assessed from intracranial EEGs (Ponten et al., 2007). The blood oxygen level-dependent (BOLD) signals used in functional MRI have been shown to be a useful tool in epileptic research (for excellent reviews, see Salek-Haddadi et al., 2003; Hamandi et al., 2004; Detre, 2006). The abnormal epileptiform activity in the cortex can be modeled with the canonical hemodynamics function of the BOLD signal, given the hypothesis that neurovascular coupling is maintained in epileptiform areas (Lemieux et al., 2008; Thornton et al., 2010). Neurovascular coupling in epileptiform tissue may vary during ictal activity, however (Salek-Haddadi et al., 2006; Zhao et al., 2009), and importantly, the BOLD signal has been shown to be altered prior to the occurrence of epileptiform activity in the EEG in addition to alterations occurring after ictal activity (Makiranta et al., 2005; Hawco et al., 2007; Jacobs et al., 2009). The BOLD changes seen prior to the EEG changes seem to be more localized than those seen following interictal spike activity (Jacobs et al., 2009). Measures that investigate the regional signal synchrony of brain BOLD signals might therefore prove to be a non-invasive source of important information on the localization and status of epileptiform activity. This would be especially useful as regards patients showing no interictal activity in routine scalp EEGs. The hemodynamic response functions used previously for detecting normal brain activation-related BOLD signal changes may be suboptimal for detecting epileptiform-related activity, and a more data-driven approach may be required (Jacobs et al., 2007). Since the neuronal synchrony of electrophysiological activity is altered in epilepsy, we suggest that the metrics of BOLD signal synchrony in the brain could reveal important information concerning the localization and status of epileptiform activity. Regional homogeneity (ReHo) analysis of BOLD signals from neighboring voxels enables the analysis of local brain activity coherence (Zang et al., 2004). This data-driven method which is suitable for exploring regional brain activity during rest involves examination of the degree of regional coherence of functional magnetic resonance imaging (fMRI) time courses. It is possible by the ReHo method to find coherent parts of active brain areas (Zang et al., 2004; Long et al., 2008). Regional homogeneity reflects the temporal homogeneity of regional BOLD signals regardless of their intensities. As the BOLD signals in fMRI reflect neural activity (Logothetis and Wandell, 2004), abnormal ReHo is probably relevant to changes in the temporal aspects of neural activity in certain brain regions, and thus ReHo may be used to detect regions with abnormal activity (Yuan et al., 2008). ReHo analysis has been used successfully to detect alterations in

subjects with ADHD, geriatric depression, schizophrenia and Alzheimer's dementia (Cao et al., 2006; Liu et al., 2006; He et al., 2007; Yuan et al., 2008). The aim here was to find out whether the synchrony of regional brain activity in resting-state fMRI, as analyzed by the ReHo method in pediatric non-lesional temporal lobe epilepsy (TLE) patients, differs from that in healthy controls, and whether it varies according to EEG findings.

2.

Results

2.1.

Patients

Twenty-one patients (11 girls and 10 boys) were examined, all with non-lesional TLE. Their mean age was 11.7 years (range 8.1–14.9) and the mean duration of epilepsy was 2.5 years. During the previous year 11 of the patients (52%) had been seizure-free and ten (48%) had one to three seizures. Interictal EEG results at the time of the examination were normal in 12 (57%) cases and abnormal in nine (43%), the abnormalities being either spikes, or sharp and/or slow waves located in the temporal lobe in all cases; on the right side in three patients, on the left side in five patients and on both sides in one patient. All the patients were receiving antiepileptic medication, 17 (81%) with one antiepileptic drug (AED) and four (19%) with two. The AEDs used were carbamazepine (cbz) in five cases, oxcarbazepine (oxc) in nine and sodium valproate (vpr) in four. The combinations of AEDs in four instances were cbz and vpr, ocx and vpr, vpr and lamotrigine, and oxc and levetiracetam.

2.2.

Controls

The control group included 21 age and gender-matched children, all with normal EEG results.

2.3.

Increased ReHo

The patients with TLE showed statistically significantly (p < 0.05, corrected) increased regional homogeneity in the left posterior cingulate gyrus, the right uncus and the right periventricular white matter (Fig. 1). When ReHo analysis was carried out in relation to the EEG findings in the epilepsy group, the patients with abnormal EEGs (subgroup EEG+, n = 9) showed significantly increased homogeneity in the right fusiform gyrus and the left medial cingulate gyrus relative to the controls (Fig. 2), while the patients with normal EEGs (subgroup EEG−, n = 12) showed significantly increased homogeneity in the left posterior cingulate gyrus and the left superior temporal gyrus, in the right inferior (BA 47) and medial frontal gyri and in the right uncus (Fig. 3).

2.4.

Decreased ReHo

Regional homogeneity was found to be decreased in the left cerebellar culmen in the epilepsy group. When ReHo analysis was carried out in relation to the EEG findings, the EEG− subgroup showed significantly decreased homogeneity in the right frontal gyrus and the left cerebellar inferior semilunar lobule (Fig. 2), while the EEG+ subgroup did not show any

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –2 29

223

Fig. 1 – Maps showing statistically significant differences between the epilepsy group and the control group (n = 21, corrected p < 0.05). Warm colors indicate increased ReHo and cold colors decreased ReHo. Talairach space coordinates are shown below the images and on the color bar with t-scores (for abbreviations see Table 1).

significant areas of decreased ReHo (Fig. 3). The statistics regarding the above areas are presented in terms of MNI coordinates in Table 1.

3.

Discussion

Four areas of altered regional homogeneity in BOLD signals were detected in the present children with non-lesional TLE during the interictal state. Importantly, the increased signal homogeneity was seen to occur over a relatively long period of time, since the functional MRI scans lasted 7.5 min. One of the areas of increased ReHo was found in the right temporal lobe. Intracortical EEG measurements have also shown increased regional synchrony of electrophysiological signals during mesial temporal lobe seizures (Ponten et al., 2007; Ortega et al., 2008). Our findings show that increased synchronization in the temporal lobe can also be detected, even in

the interictal phase, using the spatially accurate and noninvasive BOLD technique. Another significantly increased ReHo area was found in the left medio-posterior part of the cingulate gyrus in the TLE patients. The change in the upper posterior cingulate cortex was especially obvious. The posterior cingulate cortex of patients with TLE has also been shown to exhibit significant temporal BOLD signal peaks in 2-dimensional temporal cluster analysis (Morgan et al., 2007). Interestingly, our finding falls between the two regions of spike-related BOLD activation and deactivation observed in the cingulate cortex of adult patients with generalized epilepsy (Gotman et al., 2005). In particular, the area of deactivation posterior to the corpus callosum as identified by Gotman et al. (2005) was close to the area of increased regional homogeneity observed here. The cerebellar culmen showed decreased ReHo in the epilepsy group compared with the control group. Gotman et al. (2005) have shown strong positive activation correlating with

224

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –22 9

Fig. 2 – Maps showing statistically significant differences between the EEG+ subgroup and the control group (n = 9, corrected p < 0.05). Warm colors indicate increased ReHo and cold colors decreased ReHo. Talairach space coordinates are shown below the images and on the color bar with t-scores (for abbreviations see Table 1).

spike-wave EEG patterns in exactly the same area in the frontal mid-line cerebellum in their combined EEG–fMRI study of patients with generalized epilepsy, and most importantly, the source of these changes was strongly coherent EEG activity. As cerebellar function is related to fine motor control, it is plausible that the reduced homogeneity could be a sign of counterbalancing activity in an attempt to control motion. All the patients were seizure-free during the scanning, however, and no motor activity was detected. It has been shown in SPECT studies that patients with focal epilepsy have ictal hyperperfusion in the cerebellum contralaterally to the seizure focus (Bohnen et al., 1998; Dupont et al., 2009) and bilateral cerebellar hypoperfusion interictally (Van Paesschen et al., 2003; Dupont et al., 2009). Interictal hypoperfusion has been seen not only in the epileptogenic zone but also in anatomically remote areas relative to the epileptogenic focus, including the cerebellum (Van Paesschen et al., 2007). This phenomenon, called crossed cerebellar diaschisis, has been thought to indicate disconnection of the glutamatergic corticopontocerebellar tracts (Mewasingh et al., 2002; Nelissen et al., 2006). These observations together with the results of recent studies of small-world networks (see below) suggest that epilepsy is a disease of the neural networks (Spencer, 2002). The organization of the human brain cortex has recently been characterized as a scale-free small-world network (van den Heuvel et al., 2008; He et al., 2009). In scale-free networks, local sub-networks are closely interconnected by a relatively small number of hub-voxels (van den Heuvel et al., 2008). These ‘connector hubs’ have a central role in the pathophysiology of various brain diseases (Netoff et al., 2004; Achard

et al., 2006; Reijneveld et al., 2007; Morgan and Soltesz, 2008). As epileptic activity spreads in the brain it alters the network topology towards more ordered signaling with higher clustering and greater path lengths (Ponten et al., 2007). Our results show that epilepsy alters local signal synchrony at the sub-network level in the long term. Regional homogeneity analyzes the regional connectivity of the brain tissue and may be seen as a measure of the smallest network integrity measurable with fMRI techniques. It can be hypothesized that in the areas of altered homogeneity the hubs may have an abnormal synchrony signal, inducing the spread of epileptiform activity across the brain network boundaries. Further network topology investigations may reveal interesting alterations in cases of epilepsy. Interestingly, decreased ReHo signals were found in the left inferior semilunar lobule and the right inferior frontal gyrus of the patients with normal EEGs, whereas findings of increased ReHo, especially in the right hemisphere, were more prominent in this group of patients than in those with abnormal EEGs. ReHo analysis has been shown to identify white matter alterations in other diseases as well (Paakki et al., 2010). Our finding of increased ReHo in the right periventricular white matter is difficult to explain, as the BOLD signal should not receive much of a contribution from the relatively less perfused white matter. However, there have been an increasing number of preliminary findings of time domain susceptibilityweighted signal alterations, including BOLD, in white matter (Kiviniemi et al., 2009; Mezer et al., 2009), and this aspect warrants further research. Simultaneous EEG–fMRI was not performed here. It should be emphasized that all the patients had non-intractable TLE

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –2 29

Fig. 3 – Maps showing statistically significant differences between the EEG− subgroup and the control group (n = 12, corrected p < 0.05). Warm colors indicate increased ReHo and cold colors decreased ReHo. Talairach space coordinates are shown below the images and on the color bar with t-scores (for abbreviations see Table 1).

with only a few seizures and with only relatively mild interictal EEG findings or none at all. More than half of the patient group (57%) had normal interictal EEGs and the interictal EEG findings in the rest of the patients were either spikes or sharp and/or slow waves. Simultaneous fMRI-EEG studies using the ReHo method would be useful in the future

225

to obtain a further understanding of these differences and possibly explain them. Data-driven analysis of ReHo for the detection of epileptiform abnormalities avoids the problems attached to the timing of the often absent interictal activity and utilizes epilepsy-related alterations in synchrony to reveal epileptic abnormality. This is in accordance with the fact that the BOLD signal changes precede the EEG changes in focal areas (Makiranta et al., 2005; Hawco et al., 2007; Jacobs et al., 2009). Even though the EEG and BOLD signals are sampled on different spatial and temporal scales, it seems that epileptiform activity increases regional synchrony on several temporal scales that are also detectable by way of BOLD signals. The present findings suggest that the synchrony alterations observed in epilepsy are not only related to seizure onset but seem to extend to long-term regional synchrony as well (Ponten et al., 2007). This is particularly interesting, since half of our epilepsy patients had normal interictal EEGs. All our patients were receiving antiepileptic medication. To our knowledge, studies on the effect of AEDs on BOLD signal in humans have been performed only with valproate (Bell et al., 2005) and midazolam (Kiviniemi et al., 2005; Greicius et al., 2008). Valproate was shown to reduce the BOLD signal response to a stimulus and midazolam to reduce functional connectivity, probably due to the decline in blood flow at a Ramsey score 3 sedation level, where the subject responds only to strong external commands. Further studies on the effects of AEDs on functional connectivity measures such as ReHo are needed to differentiate between the effect of medication and diseaserelated alterations. The eradication of AED effects from epileptic patient data requires international collaboration efforts, and such information is beyond the scope of our current work. Such international data-base studies are nevertheless being planned in the field of resting-state brain connectome research (http:// www.nitrc.org/projects/fcon_1000/). Our results emphasize the importance of ReHo analysis as a new data-driven tool for the detection of interictal epileptiform abnormality, and possibly as an additional non-invasive tool for detecting epileptogenic foci. The results also strengthen the idea that alterations in synchronization have a central role in epileptogenesis. On the other hand, one might also argue that, due to the shortcomings of EEG, one may not be able to detect certain aspects of abnormal epileptic activity. In addition, the causal links between EEG and BOLD signals may not be understood well enough at the moment (Laufs et al., 2008). To conclude, we have shown that the regional homogeneity of BOLD signals is altered in children with non-lesional temporal lobe epilepsy even in the absence of interictal EEG abnormalities. Changes were observed both in the temporal lobe itself and in other areas, in accordance with earlier reports concerning EEG–fMRI, structural MRI and PET examinations. Our findings support the idea that the spread of altered synchronization may precede epileptiform activity.

4.

Experimental procedures

4.1.

Patients

The medical records of all children who had visited the child neurology clinics at Oulu University Hospital and Länsi-Pohja

226

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –22 9

Table 1 – Significant ReHo cluster differences between the groups (p < 0.05) in LPI oriented Talairach coordinates. Anatomical area

CM H

All patients vs. controls (n = 21) PCG L Uncus R Cerebellum, culmen L Periventricular WM R

CM Y

Mean

SEM

Max t-score

Z

Talairach coordinates MI

Volume

X

X

Y

Z

1080 972 648 405

2.4 − 37.2 0.2 − 26

34.4 1.8 51.8 25.6

48.7 − 27.5 − 6.3 26.8

3.33 3.36 − 3.18 3.20

0.08 0.13 0.09 0.12

4.66 6.33 −4.06 4.00

1.5 − 37.5 1.5 − 28.5

34.5 1.8 49.5 28.5

41.5 −27.5 −6.5 26.5

EEG+ vs. controls (n = 9) FG (RL flip) MCG

R L

783 513

− 36.9 0.9

59.7 34.5

− 11.6 42.5

3.86 4.38

0.08 0.23

5.08 7.76

− 34.5 1.5

58.5 34.5

−12.5 41.5

EEG− vs. controls (n = 12) PCG IFG (BA 47) Uncus Cerebellum, ISLL IFG (BA 46) MFG STG (BA 38)

L R R L R R L

1458 1188 1053 918 810 513 432

− 1.5 − 23.1 − 26.2 − 1.4 − 37.2 − 4.2 26.8

44.5 − 18 − 0.8 58.5 34.2 − 47 − 7.8

26.7 − 10.4 − 20.4 − 37.6 14.4 9.8 − 31.1

3.94 3.91 4.05 − 3.74 − 3.70 3.57 3.56

0.09 0.09 0.13 0.08 0.08 0.07 0.10

6.04 5.48 6.09 −4.725 −4.82 4.21 4.58

4.5 − 22.5 − 22.5 7.5 − 37.5 − 4.5 25.5

43.5 − 16.5 − 1.5 64.5 − 34.5 − 49.5 − 4.5

29.5 −12.5 −21.5 −39.5 11.5 8.5 −33.5

LPI = left posterior–inferior, H = hemisphere, volume = number of voxels per cluster, CM = center of mass, mean = mean cluster t-score, SEM = standard error of the mean, Max t-score = maximum t-score, MI = maximum intensity, PCG = posterior cingulate gyrus, WM = white matter, FG = fusiform gyrus, RL flip = right to left flipped, MCG = middle cingulate gyrus, IFG = inferior frontal gyrus, BA = Broadman area, ISLL = inferior semilunar lobule, MFG = medial frontal gyrus, STG = superior temporal gyrus.

Central Hospital between 1996 and 2007 with a diagnosis of TLE were first examined (n = 57). These hospitals are the only ones treating children with epilepsy in the area. The criteria for inclusion in the study were age between 8 and 15 years, normal 1.5 T structural MRI results and normal intelligence (intelligence quotient, IQ > 85). Brain MRI scanning is a routine examination for all epilepsy patients at the time of diagnosis. All patients with abnormal brain MRI results such as hippocampal sclerosis, dysgenetic lesions, tumors and infarcts (n = 16) were excluded. Of the 41 remaining non-lesional TLE cases, nine mentally retarded children were excluded. This meant that 32 children fulfilled our inclusion criteria and were asked to participate in the study. Ten of these did not want to participate and one did not want to continue after the first visit. The final number of patients was therefore 21. The diagnosis of TLE was based on clinical and electroencephalographic findings according to the classification of the International League Against Epilepsy (1989). In the case of a normal EEG, the TLE diagnosis was based on the clinical seizure semiology according to the ILAE criteria. The protocol included IQ testing for all participants before enrolment in order to ensure that all the patients met the requirement of a normal IQ. The study was approved by the Ethics Committee of Oulu University Hospital and informed consent was obtained from each child and at least one of the parents.

4.2.

4.3.

Protocol

The protocol involved three visits for both groups. A clinical neurological examination was performed by a pediatrician (K.M.) during the first visit and IQ was tested by a psychologist (M.H.) using the Wechsler Intelligence Scale for Children (Wechsler, 1999). If the IQ was normal (>85) and the other inclusion criteria were met, the children were then further evaluated by means of a 21-channel EEG at the second visit. The EEGs were recorded with an electrode cap and the electrodes were placed according to the international 10–10 system. The equipment used was a 40channel NicoltOne EEG system, with a sampling frequency of 256 Hz. Then recording lasted for 30–45 min and sleep recordings were obtained if feasible, the children having been deprived of sleep. The EEGs were interpreted using bipolar and common average reference montages. Functional MRI scanning was performed at the third visit. A special effort was made to prepare the children for MRI scanning in order to minimize anxiety and consequent motion artifacts. The scanning procedures were explained to the children during the first visit and pictures and a video illustrating a child in the scanning environment were shown to them. They then practiced in a tunnel-like mock scanner to experience the real feeling of lying still in a scanner, and this mock scanner training was repeated just before the real scanning. These preparatory procedures enabled scanning to be completed successfully in all cases.

Control group 4.4.

Healthy age and gender-matched controls were recruited through three state schools representing the normal Finnish child population. Children with neurological disorders, psychiatric diagnoses or known learning difficulties were excluded.

fMRI procedure

Imaging was performed using a 1.5 T General Electric Signa HDX 8-channel parallel imaging-coil ASSET system with an acceleration factor of 2.0. Hearing was protected using ear

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –2 29

plugs and motion was minimized by using soft pads fitted over the ears. T1-weighted 3D FSPGR BRAVO-sequence images (FOV 24 cm × 24 cm, 256 × 256 matrix, whole brain coverage, flip angle 20°, BW 15.63 kHz, TR 12.4 ms, TE 5.2 ms) were taken in order to obtain anatomical images for co-registration of the fMRI data with standard space coordinates. Functional scanning was performed using EPI GRE sequences (FOV 25.6 cm × 25.6 cm, 64 × 64 matrix, flip angle 90°, TR 1800 ms, TE 40 ms) with whole brain coverage using 28 oblique axial 4 mm slices with 0.4 mm spaces between the slices. The first four images were excluded owing to T1 equilibrium effects. Resting-state scanning lasted for 7 min 36 s, producing 253 brain volume data sets. The children were asked to lie still during scanning, to stay relaxed and awake and to stare at the cross on the screen.

4.5.

Pre-processing of imaging data

The fMRI data were corrected for head motion by multi-resolution rigid-body co-registration of volumes, as implemented in the FSL 3.3 MCFLIRT software (Jenkinson et al., 2002). The default settings used were: middle volume as reference, three-stage search (8 mm rough+4 mm initialized with 8 mm results+4 mm fine grain initialized with the previous 4 mm step results) with final trilinear interpolation of the voxel values and normalized spatial correlation as the optimization cost function. Brain extraction was carried out for the motion-corrected BOLD volumes by the optimization of deforming smooth surfaces, as implemented in the FSL 3.3 BET software (Smith, 2002), using the threshold

227

parameters f=0.5 and g=0, and the parameters f=0.25 and g=0 for the 3D FSPGR volumes. The BOLD data were temporally bandpass filtered (0.01
4.6.

ReHo analyses

Within-subject analysis was first performed using the ReHo approach. This was accomplished on a voxel-by-voxel basis by calculating Kendall's coefficient of concordance (KCC) (Kendall and Gibbons, 1990) of the time series of a given voxel with its 26 neighboring voxels, as described earlier (Zang et al., 2004; Yuan et al., 2008) (Fig. 4). A larger W value (between 0 and 1) for a given voxel indicates greater regional coherence within a cluster made up of the voxel and its nearest neighbors. REST (www.

Fig. 4 – Individual-level processing steps, Regional Homogeneity analyses and group level statistics, with example images for each level.

228

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –22 9

restfmri.net), a Matlab-based resting-state data analysis tool, was used to assess the individual ReHo maps. Each reading was divided by the subject's global mean KCC value within the brain mask (Cao et al., 2006; Wu et al., 2009). This is a standardization procedure similar to that used in PET studies (Raichle et al., 2001). The standardized maps were smoothed using the Gaussian kernel function (FWHM = 4 mm) for better anatomical comparability of the ReHo values at the group level. To explore the differences between the groups, paired two sample t-tests (AFNI 3dttests—paired) were performed to create group difference maps for the age and gender-matched pairs. The epilepsy group was divided into subgroups according to the interictal EEG findings at the time of the examination. Subgroups with normal interictal EEGs (EEG−, n = 12) and with interictal EEG abnormalities (EEG+, n = 9) were assessed separately against equal numbers of matched control subjects. The EEG+ subgroup was also analyzed by flipping the side of the three right temporal lobe EEG foci to the left side by means of the 3dLRflip in AFNI. Using the AFNI Monte Carlo simulation program AlphaSim (cluster connection radius 3 mm, individual voxel threshold probability 0.01, 1000 iterations), a corrected significance level of p < 0.05 was obtained for a minimum volume of 50 voxels (400 mm3) (Cox, 1996) in order to identify significant changes in ReHo. The AFNI program 3dclust was used to obtain cluster sizes, locations and their respective t-values. The probabilistic anatomical atlases included in AFNI were used to help locate the anatomical areas corresponding to clusters (Eickhoff et al., 2007). For presentation purposes, the statistical maps were transformed to Talairach coordinates (Talairach and Tournoux, 1988) and superimposed on the higher-resolution anatomical template available in MRIcro (http://www.sph.sc. edu/comd/rorden/mricro.html).

Funding This study was supported by grants from the Arvo and Lea Ylppö Foundation, the Alma and K.A. Snellman Foundation, the Foundation for Pediatric Research, the Finnish Epilepsy Association, the Maire Taponen Foundation, and Special State Grants for Health Research at the Department of Pediatrics and Adolescence, Oulu University Hospital, Finland. Support was also received from the Finnish Neurological Foundation, the Academy of Finland (grants #111711, 123772) and the Finnish Medical Foundation.

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. Bell, E.C., Wilson, M.C., Wilman, A.H., Dave, S., Silverstone, P.H., 2005. Differential effects of chronic lithium and valproate on brain activation in healthy volunteers. Hum. Psychopharmacol. 20, 415–424. Bohnen, N.I., O'Brien, T.J., Mullan, B.P., So, E.L., 1998. Cerebellar changes in partial seizures: clinical correlations of quantitative SPECT and MRI analysis. Epilepsia 39, 640–650.

Cao, Q., Zang, Y., Sun, L., Sui, M., Long, X., Zou, Q., Wang, Y., 2006. Abnormal neural activity in children with attention deficit hyperactivity disorder: a resting-state functional magnetic resonance imaging study. NeuroReport 17 (10), 1033–1036. Cox, R.W., 1996. AFNI: software for analysis and visualization of functional magnetic resonance neuroimages. Comput. Biomed. Res. 29, 162–173. Detre, J.A., 2006. Clinical applicability of functional MRI. J. Magn. Reson. Imaging 23, 808–815. Dupont, P., Zaknun, J.J., Maes, A., Tepmongkol, S., Vasquez, S., Bal, C.S., Van Paesschen, W., Carpintiero, S., Locharernkul, C., Dondi, M., 2009. Dynamic perfusion patterns in temporal lobe epilepsy. Eur. J. Nucl. Med. Mol. Imaging 36, 823–830. Eickhoff, S.B., Paus, T., Caspers, S., Grosbras, M.H., Evans, A.C., Zilles, K., Amunts, K., 2007. Assignment of functional activations to probabilistic cytoarchitectonic areas revisited. Neuroimage 36, 511–521. Gotman, J., Grova, C., Bagshaw, A., Kobayashi, E., Aghakhani, Y., Dubeau, F., 2005. Generalized epileptic discharges show thalamocortical activation and suspension of the default state of the brain. Proc. Natl. Acad. Sci. USA 102, 15236–15240. 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. Greicius, M.D., Kiviniemi, V., Tervonen, O., Vainionpää, V., Alahuhta, S., Reiss, A.L., Menon, V., 2008. Persistent default-mode network connectivity during light sedation. Hum. Brain Mapp. 29, 839–847. Hamandi, K., Salek-Haddadi, A., Fish, D.R., Lemieux, L., 2004. EEG/ Functional MRI in epilepsy: the Queen Square Experience. J. Clin. Neurophysiol. 21, 241–248. Hawco, C.S., Bagshaw, A.P., Lu, Y., Dubeau, F., Gotman, J., 2007. BOLD changes occur prior to epileptic spikes seen on scalp EEG. Neuroimage 35, 1450–1458. He, Y., Wang, L., Zang, Y., Tian, L., Zhang, X., Li, K., Jiang, T., 2007. Regional coherence changes in the early stages of Alzheimer's disease: a combined structural and resting-state functional MRI study. Neuroimage 35, 488–500. He, Y., Wang, J., Wang, L., Chen, Z.J., Yan, C., Yang, H., Hehan, T., Zhu, C., Qiyong, G., Zang, Y., Evans, A.C., 2009. Uncovering intrinsic modular organization of spontaneous brain activity in humans. PLoS ONE 4, e5226. Proposal for revised classification of epilepsies and epileptic syndromes Commission on Classification and Terminology of the International League Against Epilepsy Epilepsia 30, 389–399. Jacobs, J., Kobayashi, E., Boor, R., Muhle, H., Stephan, W., Hawco, C., Dubeau, F., Jansen, O., Stephani, Ulrich, Gotman, J., Siniatchkin, M., 2007. Hemodynamic responses to interictal epileptiform discharges in children with symptomatic epilepsy. Epilepsia 48, 2068–2078. Jacobs, J., Le Van, P., Moeller, F., Boor, R., Stephani, U., Gotman, J., Siniatchkin, M., 2009. Hemodynamic changes preceding the interictal EEG spike in patients with focal epilepsy investigated using simultaneous EEG–fMRI. Neuroimage 45, 1220–1231. Jenkinson, M., Bannister, P.R., Brady, J.M., Smith, S.M., 2002. Improved optimization for the robust and accurate linear registration and motion correction of brain images Neuroimage 17 (2), 825–841. Kendall, M., Gibbons, J.D., 1990. Rank Correlation Methods, fifth ed. Edward Arnold, New York. Kiviniemi, V.J., Haanpää, H., Kantola, J.H., Jauhiainen, J., Vainionpää, V., Alahuhta, S., Tervonen, O., 2005. Midazolam sedation increases fluctuation and synchrony of the resting brain BOLD signal. Magn. Reson. Imaging 23, 531–537. Kiviniemi, V., Starck, T., Remes, J., Xiangyu, L., Nikkinen, J., Haapea, M., Veijola, J., Moilanen, I., Isohanni, M., Zang, Y.F., Tervonen, O., 2009. Functional segmentation of the brain

BR A I N R ES E A RC H 1 3 7 3 ( 2 01 1 ) 2 2 1 –2 29

cortex using high model order group PICA. Hum. Brain Mapp. 30 (12), 3865–3886. Laufs, H., Daunizeau, J., Carmichael, D.W., Kleinschmidt, A., 2008. Recent advances in recording electrophysiological data simultaneously with magnetic resonance imaging. Neuroimage 40, 515–528. Lemieux, L., Laufs, H., Carmichael, D., Paul, J.S., Walker, M.C., Duncan, J.S., 2008. Noncanonical spike-related BOLD responses in focal epilepsy. Hum. Brain Mapp. 29, 329–345. Liu, H., Liu, Z., Liang, M., Hao, Y., Tan, L., Kuang, F., Yi, Y., Xu, L., Jiang, T., 2006. Decreased regional homogeneity in schizophrenia: a resting state functional magnetic resonance imaging study. NeuroReport 17 (1), 19–22. Logothetis, N.K., Wandell, B.A., 2004. Interpreting the BOLD signal. Annu. Rev. Physiol. 66, 735–769. Long, X.Y., Zuo, X.N., Kiviniemi, V., Yang, Y., Zou, Q.-H., Zhu, C.-Z., Jiang, T.-Z., Yang, H., Qong, Q.-Y., Wang, L., Li, K.-C., Xie, S., Zang, Y., 2008. Default mode network as revealed with multiple methods for resting-state functional MRI analysis. J. Neurosci. Meth. 171, 349–355. Lopes da Silva, F., Blanes, W., Kalitzin, S.N., Parra, J., Suffczynski, P., Velis, D.N., 2003. Epilepsies as dynamical diseases of brain systems: basic models of the transition between normal and epileptic activity. Epilepsia 44 (Suppl 12), 72–83. Lowe, M.J., Mock, B.J., Sorenson, J.A., 1998. Functional connectivity in single and multislice echoplanar imaging using resting-state fluctuations. Neuroimage 7, 119–132. Makiranta, M., Ruohonen, J., Suominen, K., Niinimäki, J., Sonkajärvi, E., Kiviniemi, V., Seppänen, T., Alahuhta, S., Jäntti, V., Tervonen, O., 2005. BOLD signal increase precedes EEG spike activity—a dynamic penicillin induced focal epilepsy in deep anesthesia. Neuroimage 27, 715–724. Matsumoto, H., Ajmone Marsan, C., 1964. Cellular mechanisms in experimental epileptic seizures. Science 144, 193–194. Mewasingh, L.D., Christiaens, F., Aeby, A., Christophe, C., Dan, B., 2002. Crossed cerebellar diaschisis secondary to refractory frontal seizures in childhood. Seizure 11, 489–493. Mezer, A., Yovel, Y., Pasternak, O., Gorfine, T., Assaf, Y., 2009. Cluster analysis of resting-state fMRI time series. Neuroimage 45, 1117–1125. Morgan, R.J., Soltesz, I., 2008. Nonrandom connectivity of the epileptic dentate gyrus predicts a major role for neuronal hubs in seizures. Proc. Natl Acad. Sci. USA 105, 6179–6184. Morgan, V.L., Gore, J.C., Abou-Khalil, B., 2007. Cluster analysis detection of functional MRI activity in temporal lobe epilepsy. Epilepsy Res. 76, 22–33. Nelissen, N., Van Paesschen, W., Baete, K., Van Laere, K., Palmini, A., Van Billoen, H., Dupont, P., 2006. Correlations of interictal FDG-PET metabolism and ictal SPECT perfusion changes in human temporal lobe epilepsy with hippocampal sclerosis. Neuroimage 32, 684–695. Netoff, T.I., Clewley, R., Arno, S., Keck, T., White, J.A., 2004. Epilepsy in small-world networks. J. Neurosci. 24, 8075–8083. Ortega, G.J., Menendez de la Prida, L., Sola, R.G., Pastor, J., 2008. Synchronization clusters of interictal activity in the lateral temporal cortex of epileptic patients: intraoperative electrocorticographic analysis. Epilepsia 49, 269–280. Paakki, J.J., Rahko, J., Long, X., Moilanen, I., Tervonen, O., Nikkinen, J., Starck, T., Remes, J., Hurtig, T., Haapsamo, H., Jussila, K., Kuusikko-Gauffin, S., Mattila, M.L., Zang, Y., Kiviniemi, V., 2010. Alterations in regional homogeneity of resting-state brain activity in autism spectrum disorders. Brain Res. 1321, 169–179.

229

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., MacLeod, A.M., Snyder, A.Z., Powers, W.J., Gusnard, D.A., Shulman, G.L., 2001. A default mode of brain function. Proc. Natl. Acad. Sci. USA 98, 676–682. Reijneveld, J.C., Ponten, S.C., Berendse, H.W., Stam, C.J., 2007. The application of graph theoretical analysis to complex networks in the brain. Clin. Neurophysiol. 118, 2317–2331. Salek-Haddadi, A., Friston, K.J., Lemieux, L., Fish, D.R., 2003. Studying spontaneous EEG activity with fMRI. Brain Res. Rev. 43, 110–133. Salek-Haddadi, A., Diehl, B., Hamandi, K., Merschhemke, M., Liston, A., Friston, K., Duncan, J.S., Fish, D.R., Lemieux, L., 2006. Hemodynamic correlates of epileptiform discharges: an EEG–fMRI study of 63 patients with focal epilepsy. Brain Res. 1088, 148–166. Smith, S.M., 2002. Fast robust automated brain extraction. Hum. Brain Mapp. 17 (3), 143–155. Spencer, S.S., 2002. Neural networks in human epilepsy: evidence of and implications for treatment. Epilepsia 43, 219–227. Talairach, J., Tournoux, P., 1988. Co-planar Stereotaxic Atlas of the Human Brain: 3-Dimensional Proportional System: An Approach to Cerebral Imaging. Thieme Medical Publishers, New York. Thornton, R.C., Rodionov, R., Laufs, H., Vulliemoz, S., Vaudano, A., Carmichael, D., Cannadathu, S., Guye, M., McEvoy, A., Lhatoo, S., Bartolomei, F., Chauvel, P., Diehl, B., De Martino, F., Elwes, R. D.C., Walker, M.C., Duncan, J.S., Lemieux, L., 2010. Imaging haemodynamic changes related to seizures: comparison of EEG-based general linear model, independent component analysis of fMRI and intracranial EEG. Neuroimage 53, 196–205. 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 Paesschen, W., Dupont, P., Van Driel, G., Van Billoen, H., Maes, A., 2003. SPECT perfusion changes during complex partial seizures in patients with hippocampal sclerosis. Brain 126, 1103–1111. Van Paesschen, W., Dupont, P., Sunaert, S., Goffin, K., Van Laere, K., 2007. The use of SPECT and PET in routine clinical practice in epilepsy. Curr. Opin. Neurol. 20, 194–202. Wechsler, D., 1999. Wechsler Intelligence Scale for Children-III. The Psychological Corporation, Sidcup UK. Wendling, F., Hernandez, A., Bellanger, J.J., Chauvel, P., Bartolomei, F., 2005. Interictal to ictal transition in human temporal lobe epilepsy: insights from a computational model of intracerebral EEG. J. Clin. Neurophysiol. 22, 343–356. Wu, T., Long, X., Zang, Y., Wang, L., Hallett, M., Li, K., Chan, P., 2009. Regional homogeneity changes in patients with Parkinson's disease. Hum. Brain Mapp. 30, 1502–1510. Yuan, Y., Zhang, Z., Bai, F., Yu, H., Shi, Y., Qian, Y., Liu, W., You, J., Zhang, X., Liu, Z., 2008. Abnormal neural activity in the patients with remitted geriatric depression: a resting-state functional magnetic resonance imaging study. J. Affect. Disord. 111 (2–3), 145–152. Zang, Y., Jiang, T., Lu, Y., He, Y., Tian, L., 2004. Regional homogeneity approach to fMRI data analysis. Neuroimage 22, 394–400. Zhao, M., Ma, Ha., Suh, M., Schwartz, T.H., 2009. Spatiotemporal dynamics of perfusion and oximetry during ictal discharges in the rat neocortex. J. Neurosci. 29, 2814–2823.