Skip to main content
NCBI home page
Human Brain Mapping logoLink to Human Brain Mapping
. 2010 Dec 22;32(11):1882–1893. doi: 10.1002/hbm.21156

Clinical functional MRI of the language domain in children with epilepsy

Marko Wilke 1,2,✉, Tom Pieper 3, Katja Lindner 1, Thekla Dushe 3, Martin Staudt 1,3, Wolfgang Grodd 4, Hans Holthausen 3, Ingeborg Krägeloh‐Mann 1
PMCID: PMC6869929  PMID: 21181799

Abstract

Functional MRI (fMRI) for the assessment of language functions is increasingly used in the diagnostic workup of patients with epilepsy. Termed “clinical fMRI,” such an approach is also feasible in children who may display specific patterns of language reorganization. This study was aimed at assessing language reorganization in pediatric epilepsy patients, using fMRI. We studied 26 pediatric epilepsy patients (median age, 13.05 years; range, 5.6–18.7 years) and 23 healthy control children (median age, 9.37 years; range, 6.2–15.4 years), using two child‐friendly fMRI tasks and adapted data‐processing streams. Overall, 81 functional series could be analyzed. Reorganization seemed to occur primarily in homotopic regions in the contralateral hemisphere, but lateralization in the frontal as well as in the temporal lobes was significantly different between patients and controls. The likelihood to find atypical language organization was significantly higher in patients. Additionally, we found significantly stronger activation in the healthy controls in a primarily passive task, suggesting a systematic confounding influence of antiepileptic medication. The presence of a focal cortical dysplasia was significantly associated with atypical language lateralization. We conclude that important confounds need to be considered and that the pattern of language reorganization may be distinct from the patterns seen in later‐onset epilepsy. Hum Brain Mapp, 2011. © 2010 Wiley‐Liss, Inc.

Keywords: functional MRI, epilepsy surgery, language reorganization, dominant hemisphere, focal cortical dysplasia, Wada test

INTRODUCTION

Functional magnetic resonance imaging (fMRI) allows for a noninvasive detection of neural activation in the brain, based on the different magnetic properties of oxy‐ and deoxyhemoglobin [Ogawa and Lee,1990]. For an examination being part of the diagnostic workup in patients, the term “clinical fMRI” was introduced [Thulborn et al.,1996]. One of the main applications of fMRI in this context has been the determination of the dominant hemisphere for language [Gaillard et al.,2004; Ruff et al.,2008; Stippich et al.,2007] for which, traditionally, the invasive Wada‐test was used [Rasmussen and Milner,1977]. As concerns about safety and ambiguous results are aggravated by its difficult application in children and/or mentally handicapped subjects, substantial effort has gone into developing fMRI as an alternative to Wada testing [Anderson et al.,2006; Gaillard et al.,2004]. Other modalities, like magnetoencephalography [MEG; Papanicolaou et al.,2004], functional transcranial doppler [fTCD; Bishop et al.,2009], or functional near‐infrared spectroscopy [fNIRS; Gallagher et al.,2008] were also used but are not yet used on a routine basis.

In this context, one seemingly obvious but not well‐researched confound is the effect of medication. This is most manifest in fMRI under sedation [Heinke et al.,2004] but must also be expected to play a role in antiepileptic drugs, which clearly have an effect on neuronal excitability. Several drugs are known to influence the observable BOLD response by reducing cerebral metabolism [Heinke et al.,2004] or blood flow [Nakao et al.,2001]. This property is particularly striking in the class of benzodiazepines, which are potent antiepileptic drugs and may “wipe out” neural activity during fMRI [Ricci et al.,2004], but a reduced BOLD response has also been seen in rodents under lamotrigine [Kida et al.,2006]. In clinical fMRI, with patients routinely being under medication, this must be expected to play a role (especially in comparison with a control group not under medication). We therefore hypothesized that an effect of medication would be present between our groups.

On the basis of the earlier studies demonstrating that early lesions to critical brain regions induce atypical language organization [Staudt et al.,2001], we also aimed at investigating whether patients with such lesions in the left frontal lobe (even remote of classical core language areas) would show a more atypical activation pattern than patients without such a lesion. As a lesion model, we decided to investigate focal cortical dysplasia (FCD). They develop late during pregnancy and local adverse events (such as an intrauterine infarct or an infection; Barkovich et al. [2001] have been linked to their emergence. As such, they may be indicative of a more severe, if not necessarily more visible, early disruption of local brain development, as also suggested by related abnormalities found in neuropathological examinations [Krsek et al.,2008]. We therefore hypothesized that the presence of a FCD would have an effect on language lateralization above and beyond that of seizure severity alone.

We here report on an ongoing study using clinically indicated functional MRI (fMRI) for the determination of hemispheric language dominance in pediatric epilepsy patients and healthy controls.

SUBJECTS AND METHODS

Patients and Control Subjects

Patients

We included all pediatric epilepsy patients referred for diagnostic workup from the Pediatric Epilepsy Center at the Behandlungszentrum Vogtareuth, Germany, in the time between January 2003 and December 2008. To be eligible, children had to demonstrate neuroradiological (lesion near suspected language regions), electrophysiological (focal EEG changes near suspected language regions), and/or clinical (impairment of productive or receptive speech during or following a seizure) evidence of their epilepsy potentially interfering with language functions. Patients were seen as part of an ongoing study to assess neural plasticity following early brain lesions [Lidzba et al.,2006; Staudt et al.,2001; Wilke et al.,2009]. Overall, we studied 26 patients (10 males and 16 females), with a median age of 156.5 months (13.05 years), range, 67.2–224.7 months (5.6–18.7 years). Median handedness as assessed with the Edinburgh Handedness Inventory [EHI; Oldfield,1971] was 0.7 (range, −1 – 1); 15 patients were right handed (range, 0.27 – 1), 8 were left handed (range, −0.64 – −1), and 3 patients were ambidextrous (range, −0.17 – 0.16). All were native German speakers. Patients needed to be able to comply with the scanning procedure and to understand the significance of the exam; therefore, only patients in the normal cognitive range or with a mild cognitive impairment were included.

Controls

Healthy controls were recruited from the community; subjects were excluded due to severe medical or neurological conditions as well as left‐handedness or a history of language problems (for more details, see Wilke et al. [2005,2006]). We included 23 controls (13 males and 10 females) with a median age of 112.4 months (9.37 years), range, 74–184.7 months (6.2–15.4 years). Handedness was right‐dominant in all controls [median EHI 0.8 (range, 0.47–1)].

General MR‐contraindications applied for all subjects (e.g., metal implants, inability to comply with the demands of an fMRI study, claustrophobia, and pregnancy). No sedation was used, and internal review board approval as well as informed parental consent and subject assent (for minors) or subject consent was obtained.

Demographic Details of Included Patients

Of the 26 patients, 10 had neuroradiological evidence of a cortical malformation, 4 suffered from tuberous sclerosis, 3 had suffered a stroke in the territory of the middle cerebral artery, and 1 each had a postoperative lesion following brain tumor removal, a Sturge–Weber phakomatosis, or a porencephalic cyst of unknown origin. The remaining six patients had no MRI abnormalities and were referred primarily for electrophysiological (EEG focus: n = 4) and for clinical reasons (speech arrest: n = 2). Suspected or confirmed lesions were localized to the parietal (n = 4), temporal (n = 4), or frontal lobe (n = 18); the right side was affected in two, the left side in 24 patients. Demographic details of all patients are listed in Table I.

Table I.

Demographic details of the patients included in this study

ID Age (Y) Gender (M/F) EHI MRI Loc (P/F/T) Side (L/R) Age at onset (Y) Seizure type (F/M/SG) Speech arrest (Y/N) Current AED Failed AED Seizure frequency
P01 14.5 M 1 (R) FCD P R 6 F N OXC, LTG 6 Daily
P02 12.5 F 1 (R) Normal T L 5 F N LTG 4 Monthly
P03 5.7 F −0.64 (L) TS F L 0.25 F N OXC, LEV, VGB, CLB 10 Daily
P04 15.9 F 0.8 (R) Tu F L 14 F Y LEV 3 Daily
P05 11.2 M −0.11 (AD) Malformation F L 7 M N VPA, LTG 8 Weekly
P06 17.5 F −1 (L) TS F L 1.5 M Y OXC 6 Daily
P07 14.4 F 0.67 (R) Normal F L 10 M Y LTG 5 Weekly
P08 7.5 M 1 (R) FCD F L 4 M N OXC, LTG 9 Daily
P09 16.7 M 0.27 (R) FCD F L 2 SG Y LEV, LTG 7 Monthly
P10 10.6 F 0.82 (R) TS F L 2 M N VPA, OXC, LTG 7 Weekly
P11 18.4 M 0.87 (R) FCD P L 10 SG N OXC, LTG 5 Monthly
P12 17.7 M −0.17 (AD) FCD F L 2.5 M N VPA 14 Daily
P13 14.9 F 0.67 (R) FCD T L 5.5 F N OXC, LEV 6 Daily
P14 5.8 M 1 (R) FCD F L 1.25 F N OXC 3 Daily
P15 18.7 F −0.75 (L) SWS F L 0.5 SG N OXC, LTG 6 Monthly
P16 14.7 F 1 (R) Porencephalic cyst P L 14 SG N OXC 1 Less
P17 12.0 F −1 (L) Infarct F L 10 F N OXC 3 Monthly
P18 7.9 F 1 (R) FCD T L 2 M N OXC, LEV 5 Daily
P19 14.5 F −0.73 (L) Infarct P L 11 SG N VPA, OXC 2 Monthly
P20 9.6 F 0.67 (R) Normal F L 7 F N VPA, OXC 8 Daily
P21 5.6 F −1 (L) Infarct F L 3 F Y OXC, VPA 2 Monthly
P22 9.9 F 1 (R) Normal F L 7 SG N VPA, OXC 3 Daily
P23 12.5 F −0.67 (L) Normal F L 4 M N OXC 4 Less
P24 17.5 M 0.17 (AD) TS T R 2 SG N OXC, VPA 4 Monthly
P25 7.9 F −1 (L) FCD F L 0.33 M N LTG, VPA 8 Daily
P26 13.6 M 1 (R) Normal F L 0.75 M Y VPA, LTG 6 Monthly
Open in a new tab

M/F, male, female; L/R/AD, left, right, ambidextrous; FCD, focal cortical dysplasia; TS, tuberous sclerosis; Tu, tumor; SWS, Sturge–Weber syndrome; P/F/T, parietal, temporal, frontal lobe; F/M/SG, focal, mixed, secondarily generalized; AED, antiepileptic drug. See text for more details.

Seizures were classified as focal in 9 and mixed in 11 patients; secondary generalized seizures were seen in six patients. Seizure frequency was daily in 12 patients, weekly in 3 patients, monthly in 9 patients, and less than 1 per month in 2 patients. Median age of seizure onset was 4 years (range, 0.25–14), and median duration of epilepsy was 6.5 years (range, 1–18). All patients were under medication, with a median of 2 (range, 1–4) antiepileptic drugs at the time of scan. Most frequently used were oxcarbazepine (OXC; n = 18), lamotrigine (LTG; n = 11), valproate (VPA; n = 10), and leviracetam (LEV; n = 5); other medications included vigabatrine and clobazam (each n = 1). The patients had failed a median of six medications (range, 1–14) in the past.

fMRI Tasks

All tasks were previously established for use in young children [Wilke et al.,2005,2006]. The decision on which tasks to use was based on the evidence of language impairment. More than one task was performed whenever possible [Gaillard et al.,2004].

The beep‐stories task was used to tap receptive language functions and, thus, language processing primarily in the temporal lobe. In this task, simple 30‐s‐long children's stories were modified, so that four to six key words in each story were replaced with a sinus tone (200 Hz and 750 ms). As these words are necessary to follow the story, the subject has to “think harder,” inducing an increase of inferior‐frontal brain activation [Wilke et al.,2005]. This task is especially suited for younger subjects [Wilke et al.,2005].

To assess productive language functions, and thus, language processing primarily in the frontal lobe, the letter‐task was used: this task consists of the image of an object, upon which the subject had to decide if the sound of the vowel “I” (always pronounced [ee] in German) is present in the name of this object. In the control condition, two abstract and unnamable images were presented, and the subject has to decide if one fits “like the piece of a puzzle” into the other one. This task activates inferior‐frontal language regions and is doable by children as young as 6 years of age [Wilke et al.,2006].

MRI Scanning and Data Processing

All subjects were scanned on the same 1.5 T Siemens Sonata MR‐scanner (Siemens Medizintechnik, Erlangen, Germany) using a standard quadrature head coil. An EPI‐sequence was used to acquire functional series (TR = 3,000 ms, TE = 40 ms, 40 axial, 3‐mm slices with 0.5‐mm gap, in‐plane matrix = 64 × 64, voxel size 3 × 3 × 3.5 mm3). A T1‐weighted anatomical 3D‐dataset (176 contiguous sagittal slices, in‐plane matrix 256 × 256, voxel size 1 × 1 × 1 mm3) and a gradient‐echo B0‐field map (with the same voxel size and slice prescription as in the functional series) were also acquired.

FMRI data were processed and analyzed using spm5 (Wellcome Department of Imaging Neuroscience, University College, London, UK) running in Matlab (MathWorks, Natick, MA). The first 10 volumes (plus two prescans) were removed to allow for stabilization of longitudinal magnetization, leaving 100 EPI volumes for analyses (five blocks each of the active and control condition). Initially, images were subjected to a wavelet‐based denoising [Wink and Roerdink,2004], followed by the removal of EPI‐distortions and B0* movement interaction effects using the individual field map [Andersson et al.,2001]. Simultaneously, data were motion‐corrected, and series with motion on any frame exceeding voxel size (3 mm) were discarded. Normalization was achieved by segmenting the anatomical images [Ashburner and Friston,2005] based on custom‐made pediatric reference information [Wilke et al.,2008] to avoid processing bias [Wilke et al.,2002]. Following coregistration, parameters were then applied to the functional series (final resolution 3 × 3 × 3 mm). Global image signal drifts were removed [Macey et al.,2004], and the data was smoothed using a 12‐mm (FWHM) isotropic Gaussian filter.

Data Analysis

Statistical analysis on the single‐subject level was performed applying the framework of the general linear model [Friston et al.,1995], using a box‐car reference function convolved with the hemodynamic response function. For the illustration of group activation patterns on the second level, we used a nonparametrical statistical approach, using SnPM3 on SPM5 owing to group size considerations. Subject gender and age (in months) were used as covariates of no interest. Variance smoothing of FWHM = 8 mm was used to increase sensitivity by pooling the variance estimates, and significance was assumed at P ≤ 0.05, FWE‐corrected for multiple comparisons, using the combined cluster/voxel information [Hayasaka and Nichols,2004].

Lateralization was assessed by calculating a lateralization index according to LI = (L − R)/(L + R), using the sum of voxel values from the individual t‐maps. We used a dedicated toolbox providing standard masks [Wilke and Lidzba,2007]. On the basis of the previous studies [Gaillard et al.,2004; Liégeois et al.,2004; Mbwana et al.,2009; Wilke et al.,2005,2006], we decided to investigate the temporal and the frontal lobe. In the case of obvious destructive lesions (tumors, cysts, and stroke), results from the affected lobe were not used. We used a bootstrapping approach [Wilke and Schmithorst,2006] with default settings (no optional steps, bootstrap sample size: 25%, minimum bootstrap sample size: five voxels, maximum bootstrap sample size: 10,000 voxels). As in other studies [Rosenberger et al.,2009; Ruff et al.,2008; Wilke et al.,2005,2006], values of −0.2 < LI < 0.2 were considered bilateral (B); values < −0.2 were considered right‐dominant (R), values > 0.2 were considered left‐dominant (L; with ±1 being exclusively right/left).

To investigate a potential effect of medication, local activation strength was determined by assessing the parameter estimates of the active when compared with the control condition [Gizewski et al.,2007]. To set spatial specificity, nonparametrical second‐level analyses (for the beep‐stories task, see above) were calculated, and the weighted center of mass [Wilke et al.,2009] of the overlapping activation was determined. Relative signal change was determined for each subject from this and all surrounding voxels (i.e., 27 voxels). As no attempt was made to scale the values, results are given in arbitrary units.

To further assess the influence of FCD on atypical language organization, we investigated a subsample of 10 patients with a left inferior‐frontal EEG focus. Of these patients, five (FCD−: P07, P20, P22, P23, and P26) had MRIs repeatedly read as normal. The other five (FCD+: P08, P9, P12, P14, and P25) had neuroradiological evidence of a FCD in the left frontal lobe. In all cases, this was not within a putative Broca's area (insula, orbital gyrus, and rectal gyrus in one patient each; medial frontal gyrus in two patients). The two patient groups (FCD− and FCD+) did not differ significantly with regard to age, age at seizure onset, duration of disease, handedness, gender composition, current or past number of antiepileptic drugs, or seizure frequency. Of our tasks, only the beep‐stories task was completed by all 10 patients.

Because of partly small numbers and potentially unequal variance, we decided to use robust nonparametrical tests to assess significance between demographic variables. Differences in continuous variables were assessed using the Mann–Whitney U test, whereas categorical variables were assessed using the χ2 test. Correlations were assessed using Kendall's rank correlation, and differences in variance between patients and controls were assessed using Levene's test [Van Halen, 2005]. Significance was assumed at P ≤ 0.05.

RESULTS

Subjects

Patients were significantly older than healthy controls (median age, 13.05 vs. 9.37 years, P = 0.02). Gender composition was not significantly different. Comparing the EHI‐scores showed no significant differences, but there were significantly more left‐handed and ambidextrous subjects in the patient group (P < 0.001).

Functional Series

Overall, 48 series of patients (beep‐stories, n = 26; letter‐task, n = 22) and 33 series of healthy controls (beep‐stories, n = 17; letter‐task, n = 16) could be analyzed. In these series, median maximum motion was 0.64 mm (range, 0.2–2.9 mm) and 0.49 mm (range, 0.19–1.5 mm), respectively, in the patients and 0.56 mm (range, 0.13–3 mm) and 0.81 mm (range, 0.14–2.4 mm), respectively, in the controls (not significant).

Activation Strength (Beep Stories Task)

The strength of activation in patients and healthy controls is shown in Figure 1. For both regions, healthy controls showed significantly stronger activation (P = 0.03 and P = 0.005, respectively).

Figure 1.

Figure 1

Open in a new tab

Activation strength in the beep‐stories task: following group analyses of the controls (upper panel, left) and the patients (upper panel, right), the centers of overlapping activations were determined (upper panel, middle). Within both these regions, the healthy controls showed significantly stronger activation (P = 0.03 [L] and 0.005 [R]). MR images are in neurological orientation, and results are overlaid on the custom‐made pediatric gray matter map. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

fMRI Results

Group activation maps and lateralization indices from the four tasks are shown in Figures 2 and 3. Significant differences in lateralization were found for the frontal lobe (P = 0.02) and the temporal lobe (P = 0.009) in the beep‐stories task (see Fig. 2). The differences in the letter‐task did not reach significance (see Fig. 3).

Figure 2.

Figure 2

Open in a new tab

Activation patterns and lateralization indices in the beep‐stories task: activation patterns of controls (upper panels, left and right) and patients (lower panels, left and right), and box‐Whisker‐plots of the lateralization indices in the frontal (F) and the temporal (T) lobe. Within both lobes, lateralization was significantly different between patients and controls (P = 0.02 [F] and 0.009 [T]). Results are rendered on the 3D surface of the custom‐made gray‐matter map. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3.

Figure 3

Open in a new tab

Activation patterns and lateralization indices in the letter task: activation patterns of controls (upper panels, left and right) and patients (lower panels, left and right), and box‐Whisker‐plots of the lateralization indices in the frontal (F) and the temporal (T) lobe. Within both lobes, the variance of lateralization indices was significantly different between patients and controls (P = 0.01 [F] and 0.02 [T]). Results are rendered on the 3D surface of the custom‐made gray matter map. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Regarding the presence of typical versus atypical language lateralization, there were significant differences in the frontal lobe (P = 0.01) and the temporal lobe (P = 0.02) in the beep‐stories task as well as in the temporal lobe in the letter‐task (P = 0.01). For more details, see Table II.

Table II.

Results of categorization w.r.t. language dominance

Task Lobe All patients Controls χ2 test Patients (FCD−) χ2 testa Patients (FCD+) χ2 testb
L B R L B R L B R L B R
Beep stories Frontal lobe 6 7 11 12 3 2 P = 0.01 3 1 1 n.s. 0 1 4 P = 0.007
Temporal lobe 6 1 19 9 3 5 P = 0.02 3 0 2 n.s. 0 1 4 P = 0.05
Letter task Frontal lobe 13 3 4 14 2 0 n.s.
Temporal lobe 14 1 7 14 2 0 P = 0.01
Open in a new tab

L, left‐dominant; B, bilateral; R, right‐dominant; n.s., not significant; FCD−, patients without evidence for a focal cortical dysplasia; FCD+, patients with evidence for a focal cortical dysplasia.

a

Patients (FCD−) versus Controls.

b

Patients (FCD+) versus Controls; see text for more details.

Regarding the variance of the resulting lateralization indices as assessed by Levene's test, these were significantly different between patients and controls in the frontal (P = 0.01) and temporal (P = 0.02) lobe in the letter‐task.

Regarding the effect of left‐frontal FCD on language lateralization in the beep‐stories task, only FCD+ patients were significantly more right‐lateralized than controls in both frontal (P = 0.006) and temporal lobe (P = 0.02; Fig. 4). Regarding the presence of atypical language, only FCD+ patients were significantly more likely to show atypical language lateralization in both frontal (P = 0.007) and temporal lobe (P = 0.05; Table II).

Figure 4.

Figure 4

Open in a new tab

Subgroup comparison of the lateralization indices from the beep‐stories task: lateralization in the frontal (F) and temporal (T) lobe in healthy controls (C) and patients without (FCD−) and with (FCD+) evidence of a left‐frontal focal cortical dysplasia. Black bars indicate median value in each group. Results are not significantly different between controls and FCD− but are significantly different between controls and FCD+, in the frontal (P = 0.006) and the temporal lobe (P = 0.02). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Influence of Demographic Details

Seizure type, age at onset or duration of epilepsy, presence of speech arrest, number of current, or past antiepileptic drugs were not correlated with lateralization in any task. Seizure frequency was correlated (P = 0.03) with lateralization in the temporal lobe in the letter‐task (more frequent seizures were associated with stronger right‐lateralization). EHI‐scores were significantly correlated with lateralization in the temporal lobe in the beep‐stories (P = 0.001) and the frontal lobe in the letter‐task (P = 0.03; more left‐lateralized handedness was associated with stronger right‐lateralization).

DISCUSSION

In this work, we report on different aspects of performing clinically indicated fMRI in children as part of the presurgical assessment in children with epilepsy. It should first be noted that there are specific aspects to performing fMRI studies in children. On one hand, fMRI even in healthy children already poses a number of special challenges: from a child‐friendly task design and extensive subject preparation to pediatric particularities during data processing, great care must be taken to appropriately account for the special needs of children [O'Shaughnessy et al.,2008; Wilke et al.,2002,2006,2008]. This is all the more true for children with a neurological disorder like epilepsy: both the disorder itself and the necessary medication may influence the child's ability to cooperate [Yerys et al.,2009]. On the other hand, children may be candidates for surgical approaches not feasible in adults, and the possible gain of function may be worth taking risks not justifiable in an adult subject [Bast et al.,2006; Jonas et al.,2004]. This context makes efforts to establish clinically indicated language fMRI in children all the more pressing.

Confounding Effects of Antiepileptic Drugs

To test the hypothesis of a medication effect, we assessed activation strength in the temporal lobe in one task (beep‐stories). In this primarily passive paradigm, temporal activation should not be a function of attention or effort; also, there was no significant difference in motion, age, or gender between the patients and controls performing this task; these factors can therefore be ruled out as relevant confounders. Although it should be noted that seizure severity itself has been linked to suppressing cortical activation [Jayakar et al.,2002], the reduction in the BOLD response was not correlated with seizure frequency in our patients, making this a less‐likely confound. As can be seen from Figure 1, there is indeed a significant difference in the strength of the activation between the groups, with healthy controls showing significantly stronger activation in both hemispheres than epilepsy patients. The fact that this effect is present in both hemispheres argues against an effect of different lateralization in the temporal lobe (see below). Owing to group‐size limitations and additive effects of more than one drug, no more detailed analyses on the effects of single medications could be performed (although it seems interesting to note that only one of our patients was under medication with a benzodiazepine). It is too early to speculate on whether these are direct or indirect effects [Lowen et al.,2009] of the antiepileptic drugs, although the former seems much more likely: any medication aimed at avoiding epileptic discharges must be expected to also influence perisynaptic activity, considered to underlie the detectable BOLD response [Logothetis and Pfeuffer,2004].

We believe that our results support the notion that antiepileptic medication may be an important confound when performing imaging studies in patient populations. Such a systematic group effect of course has several important repercussions. For one, when generating results that show a strong threshold‐dependency, such as a lateralization index [Ruff et al.,2008; Wilke and Lidzba,2007], this threshold has to consider the potentially lower level of activation in patients under medication. In our case, a threshold‐free, weighted approach to calculating a lateralization index [Wilke and Schmithorst,2006] takes this into account as no single threshold of significance was used. This effect may also contribute to previous findings demonstrating that not one threshold is applicable for both patients and controls [Rosenberger et al.,2009; Ruff et al.,2008]. Moreover, a direct statistical comparison between activation patterns of patients and controls implicitly assumes similar levels of activation. If this is not the case, quantitative, not qualitative, differences in the pattern of activation will be detected. For this reason, we here also abstained from directly contrasting the activation patterns of patients versus controls in the same statistical analysis. We suggest that, in the presence of such a systematic confound, individually thresholded analyses may be more sensitive in detecting true group differences. In the future, a multimodal assessment may aid to further elucidate the factors driving this effect, including combined EEG‐fMRI studies [Moeller et al.,2009] and PET studies assessing energy metabolism (where both hypo‐ and hypermetabolic interictal states have been described; [Chugani et al.,1990; Maquet et al.,1995]).

Group Effects and Statistical Considerations

Hemispheric dominance for language has been of major interest to neuroscience, and the hope to avoid (or at least, predict) postsurgical deficits has fueled attempts to firmly establish clinical fMRI [Ruff et al.,2008; Stippich et al.,2007; Thulborn et al.,1996]. Group activation maps are of limited use when assessing a diverse population, and our analyses of lateralization show how different the patients are compared to the healthy controls: within our patient group, the range of detected lateralization indices was much wider in both tasks (Figs. 2 and 3). Consequently, the likelihood to detect atypical language organization was higher in the patient group (see Table II for details), underlining the value of assessing language in these selected patients. Of note, this greater range of individual results is not necessarily apparent in the group‐activation maps included for illustration: although the beep‐stories (see Fig. 2) show apparent differences between patients and controls in the group analyses especially in the (“mirrored”) frontal lobe activation, the activation patterns in the letter‐task seem rather similar (see Fig. 3). Still, the range of observed lateralization indices is much wider in the patients than in the controls in all tasks, confirming the limited informative value of group activation maps. In the letter task, the variance in the patients was significantly larger than in the controls, which is an important confound for statistical comparisons [Van Halen, 2005] and was taken into account here by using nonparametric statistical approaches throughout. For statistical reasons [Mbwana et al.,2009; Mühlau et al.,2009], it is difficult to compare a single individual against a group of controls. Approaches based on several patient‐control comparisons [Mühlau et al.,2009] or |z|‐score maps [Mbwana et al.,2009] have recently been proposed, but there is no general consensus on how to calculate such comparisons. Thus, it is still difficult to objectively determine, in each individual patient, how abnormal an activation pattern actually is. As a group, however, the much higher prevalence of atypical results is clearly evident in our sample, confirming previous studies [Mbwana et al.,2009; Möddel et al.,2009; Rosenberger et al.,2009].

Regarding the influence of demographic variables on lateralization in the patient group, we found a significant influence of handedness on lateralization in the temporal (beep‐stories tasks) as well as the frontal lobe (letter‐task). This is well in line with a more variable activation pattern in left‐handed healthy subjects [Szaflarski et al.,2002] as well as in epilepsy patients [Mbwana et al.,2009; Möddel et al.,2009]. It is of course important to remember that right‐lateralized language and left‐handedness in patients may both be due to a common adverse factor affecting the left hemisphere. This, however, is difficult to investigate in our sample as left‐handedness was already considered an additional risk factor, resulting in a referral bias. We found no further significant influence of other demographic variables (such as age at seizure onset or duration of epilepsy) on language lateralization, in line with recent results [Rosenberger et al.,2009].

Specific Effects

It was previously shown that, in the case of nonepileptic early brain lesions, language may be reorganized into homotopic brain regions in the opposite hemisphere [Lidzba et al.,2006; Staudt et al.,2001]. However, in the case of epilepsy, the picture is less clear. There is now good evidence that epilepsy may cause changes in the organization of language within the brain [Mbwana et al.,2009; Rosenberger et al.,2009; Voets et al.,2006], which may even be partially reversible [Helmstaedter et al.,2006]. However, several authors have not found an exclusive language reorganization into right‐hemispheric homotopic brain regions [Liégeois et al.,2004; Mbwana et al.,2009; Voets et al.,2006], indicating that the diverse impact of epileptic activity may also lead to a more diverse pattern of functional reorganization.

Qualitatively comparing the group activation maps of our two tasks shows that, as a whole, the patients activate homotopic right‐hemispheric regions, which is particularly obvious in the frontal lobe activation in the beep‐stories task (see Fig. 1). As shown before in the case of early lesions [Staudt et al.,2001], the activation pattern in the patients looks “like a mirror image” of the pattern observable in healthy controls. Although we did not specifically investigate distinct topographical differences below the lobar level between the groups, these findings indicate that interhemispheric, homotopic reorganization of language functions does seem to be present not only in the case of early lesions [Staudt et al.,2001] but also in the case of early‐onset epilepsy, confirming previous results [Rosenberger et al.,2009]. The limits to this pattern of reorganization were recently shown in adults with late‐onset epilepsy [Mbwana et al.,2009] where intra‐hemispheric reorganization was observed in a subset of patients. Whether age or other factors associated with epilepsy are the decisive factors explaining these effects can only be investigated in more detail in a sample comprising patients with early‐ as well as late‐onset epilepsy.

An interesting question is whether the nature of the lesion inducing epilepsy is an important factor when it comes to language reorganization. In temporal lobe epilepsy, no differences were found for early or late lesion types [Briellmann et al.,2006], nor was an influence of pathology found in a recent study in left‐hemispheric focal epilepsy [Rosenberger et al.,2009]. We decided to further investigate this question in our sample comparing patients with a left‐frontal EEG focus who had either a FCD+ or were MRI‐negative (FCD−). FCDs are common, early malformations of cortical development, and are frequently associated with epilepsy [Bast et al.,2006; Fauser et al.,2006], but their classification is a matter of ongoing debate [Barkovich et al.,2001; Krsek et al.,2008]. They are found in a significant number of patients undergoing epilepsy surgery and compared to histological results, MRI may be both false‐negative and false positive [Fauser et al.,2006; Krsek et al.,2008; Widdess‐Walsh et al.,2006]. In our sample, they have a clear effect on language lateralization: the presence of a FCD in the left frontal lobe was associated with a significantly more atypical language lateralization compared to controls (see Fig. 4). In other words, the finding of a left‐frontal FCD made atypical language organization much more likely, even though it was remote from classical language regions. This is of potentially high‐clinical relevance and would be in line with previous studies where even small, early brain lesions remote from Broca's area led to a shift in hemispheric dominance [Staudt et al.,2001]. This argument is also strongly supported by results from a recent study using the Wada test where it was suggested that an early left‐sided lesion is most likely to result in right‐hemispheric dominance [Möddel et al.,2009]. The discrepancy with earlier (negative) findings may be explained by the fact that we were able to only assess left‐frontal FCD as opposed to temporal [Briellmann et al.,2006] or, even more general, left‐hemispheric foci [Rosenberger et al.,2009]. If confirmed, left‐frontal malformations of cortical development could thus constitute a second model lesion leading to atypical language organization.

Limitations of this Study

Clearly, our study has several limitations, mainly the ultimately small number of subjects in the different subgroups. For example, more detailed analyses of the effect of different medications on the activation strength are not possible, given the number of patients. It should also be noted that only longitudinal data would allow to unequivocally attribute the effect seen in our patients to their medication at the time of scan. Moreover, our patients were slightly (and, as a whole, significantly) older than our controls, which is a potential confound. However, as age increases both the amplitude of the BOLD response [Schapiro et al.,2004] and lateralization to the left in different language tasks [Everts et al.,2009; Holland et al.,2001], the fact that we observed a weaker BOLD response and a less‐lateralized pattern in our (older) patients only underlines the robustness of our results. We could not reproduce the findings from the FCD± patients as the letter task was completed by all patients in the FCD− group, but only by three patients in the FCD+ group, which is prohibitively small for a group comparison. The former group did not differ significantly from the control population (median LI = 0.73 vs. 0.635 [frontal] and 0.52 vs. 0.59 [temporal], in patients and controls, respectively; all P ≫ 0.05). However, as the latter patients (FCD+) are the decisive patient group and as there are only three data points available (median LI = 0.62 [frontal] and 0.39 [temporal]), no meaningful statistic can be calculated. Although not enough data are available on the postsurgical outcome in those patients that since were operated upon, this was not the focus of this study. Moreover, neuropsychological evaluation was custom‐tailored for each patient and thereby too inhomogeneous (and not always done in close proximity to the MRI scan) to allow further analyses. Therefore, further research in larger groups seems necessary to explore these specific issues.

To conclude, clinically indicated language fMRI is feasible in children with epilepsy. An effect of antiepileptic medication on the observable activation strength seems to be present and, if confirmed, would need to be taken into account when comparing patients and healthy controls. Atypical language organization is much more common in pediatric epilepsy patients and even more so in left‐handed patients or in patients with a left‐frontal FCD. Language reorganization in children with epilepsy seems to take place predominantly in homotopic brain regions in the opposite hemisphere. These factors should be taken into account when using fMRI in the context of planning neurosurgical interventions for seizure relief in children.

Acknowledgements

We thank all the patients and the control subjects for participating in this study. None of the authors has any conflict of interest to disclose.

REFERENCES

  1. Anderson DP, Harvey AS, Saling MM, Anderson V, Kean M, Abbott DF, Wellard RM, Jackson GD ( 2006): FMRI lateralization of expressive language in children with cerebral lesions. Epilepsia 47: 998–1008. [DOI] [PubMed] [Google Scholar]
  2. Andersson JL, Hutton C, Ashburner J, Turner R, Friston K ( 2001): Modeling geometric deformations in EPI time series. Neuroimage 13: 903–919. [DOI] [PubMed] [Google Scholar]
  3. Ashburner J, Friston KJ ( 2005): Unified segmentation. NeuroImage 26: 839–851. [DOI] [PubMed] [Google Scholar]
  4. Barkovich AJ, Kuzniecky RI, Jackson GD, Guerrini R, Dobyns WB ( 2001): Classification system for malformations of cortical development: Update 2001. Neurology 57: 2168–2178. [DOI] [PubMed] [Google Scholar]
  5. Bast T, Ramantani G, Seitz A, Rating D ( 2006): Focal cortical dysplasia: Prevalence, clinical presentation and epilepsy in children and adults. Acta Neurol Scand 113: 72–81. [DOI] [PubMed] [Google Scholar]
  6. Bishop DV, Watt H, Papadatou‐Pastou M ( 2009): An efficient and reliable method for measuring cerebral lateralization during speech with functional transcranial Doppler ultrasound. Neuropsychologia 47: 587–590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Briellmann RS, Labate A, Harvey AS, Saling MM, Sveller C, Lillywhite L, Abbott DF, Jackson GD ( 2006): Is language lateralization in temporal lobe epilepsy patients related to the nature of the epileptogenic lesion? Epilepsia 47: 916–920. [DOI] [PubMed] [Google Scholar]
  8. Chugani HT, Shields WD, Shewmon DA, Olson DM, Phelps ME, Peacock WJ ( 2006): Infantile spasms. I. PET identifies focal cortical dysgenesis in cryptogenic cases for surgical treatment. Ann Neurol 27: 406–413. [DOI] [PubMed] [Google Scholar]
  9. Everts R, Lidzba K, Wilke M, Kiefer C, Mordasini M, Schroth G, Perrig W, Steinlin M ( 2009): Strengthening of laterality of verbal and visuospatial functions during childhood and adolescence. Hum Brain Mapp 30: 473–483. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Fauser S, Huppertz HJ, Bast T, Strobl K, Pantazis G, Altenmueller DM, Feil B, Rona S, Kurth C, Rating D, Korinthenberg R, Steinhoff BJ, Volk B, Schulze‐Bonhage A ( 2006): Clinical characteristics in focal cortical dysplasia: A retrospective evaluation in a series of 120 patients. Brain 129: 1907–1916. [DOI] [PubMed] [Google Scholar]
  11. Fernández G, Specht K, Weis S, Tendolkar I, Reuber M, Fell J, Klaver P, Ruhlmann J, Reul J, Elger CE ( 2003): Intrasubject reproducibility of presurgical language lateralization and mapping using fMRI. Neurology 60: 969–975. [DOI] [PubMed] [Google Scholar]
  12. Friston KJ, Holmes AP, Worsley KJ, Poline JB, Frith C, Frackowiak RSJ ( 1995): Statistical parametric maps in functional imaging: A general linear approach. Hum Brain Mapp 2: 189–210. [Google Scholar]
  13. Gaillard WD, Balsamo L, Xu B, McKinney C, Papero PH, Weinstein S, Conry J, Pearl PL, Sachs B, Sato S, Vezina LG, Frattali C, Theodore WH ( 2004): fMRI language task panel improves determination of language dominance. Neurology 63: 1403–1408. [DOI] [PubMed] [Google Scholar]
  14. Gallagher A, Bastien D, Pelletier I, Vannasing P, Legatt AD, Moshé SL, Jehle R, Carmant L, Lepore F, Béland R, Lassonde M ( 2008): A noninvasive, presurgical expressive and receptive language investigation in a 9‐year‐old epileptic boy using near‐infrared spectroscopy. Epilepsy Behav 12: 340–346. [DOI] [PubMed] [Google Scholar]
  15. Gizewski ER, de Greiff A, Maderwald S, Timmann D, Forsting M, Ladd ME ( 2007): fMRI at 7 T: Whole‐brain coverage and signal advantages even infratentorially? NeuroImage 37: 761–768. [DOI] [PubMed] [Google Scholar]
  16. Hayasaka S, Nichols TE ( 2004): Combining voxel intensity and cluster extent with permutation test framework. NeuroImage 23: 54–63. [DOI] [PubMed] [Google Scholar]
  17. Heinke W, Fiebach CJ, Schwarzbauer C, Meyer M, Olthoff D, Alter K ( 2004): Sequential effects of propofol on functional brain activation induced by auditory language processing: An event‐related functional magnetic resonance imaging study. Br J Anaesth 92: 641–650. [DOI] [PubMed] [Google Scholar]
  18. Helmstaedter C, Fritz NE, González Pérez PA, Elger CE, Weber B ( 2006): Shift‐back of right into left hemisphere language dominance after control of epileptic seizures: Evidence for epilepsy driven functional cerebral organization. Epilepsy Res 70: 257–262. [DOI] [PubMed] [Google Scholar]
  19. Holland SK, Plante E, Weber Byars A, Strawsburg RH, Schmithorst VJ, Ball WS ( 2001): Normal fMRI brain activation patterns in children performing a verb generation task. NeuroImage 14: 837–843. [DOI] [PubMed] [Google Scholar]
  20. Jayakar P, Bernal B, Santiago Medina L, Altman N ( 2002): False lateralization of language cortex on functional MRI after a cluster of focal seizures. Neurology 58: 490–492. [DOI] [PubMed] [Google Scholar]
  21. Jonas R, Nguyen S, Hu B, Asarnow RF, LoPresti C, Curtiss S, de Bode S, Yudovin S, Shields WD, Vinters HV, Mathern GW ( 2004): Cerebral hemispherectomy: Hospital course, seizure, developmental, language, and motor outcomes. Neurology 62: 1712–1721. [DOI] [PubMed] [Google Scholar]
  22. Kida I, Smith AJ, Blumenfeld H, Behar KL, Hyder F ( 2006): Lamotrigine suppresses neurophysiological responses to somatosensory stimulation in the rodent. NeuroImage 29: 216–224. [DOI] [PubMed] [Google Scholar]
  23. Krsek P, Maton B, Korman B, Pacheco‐Jacome E, Jayakar P, Dunoyer C, Rey G, Morrison G, Ragheb J, Vinters HV, Resnick T, Duchowny M ( 2008): Different features of histopathological subtypes of pediatric focal cortical dysplasia. Ann Neurol 63: 758–769. [DOI] [PubMed] [Google Scholar]
  24. Lidzba K, Staudt M, Wilke M, Krägeloh‐Mann I ( 2006): Visuospatial deficits in patients with early left‐hemispheric lesions and functional reorganization of language: Consequence of lesion or reorganization? Neuropsychologia 44: 1088–1094. [DOI] [PubMed] [Google Scholar]
  25. Liégeois F, Connelly A, Cross JH, Boyd SG, Gadian DG, Vargha‐Khadem F, Baldeweg T ( 2004): Language reorganization in children with early‐onset lesions of the left hemisphere: An fMRI study. Brain 127: 1229–1236. [DOI] [PubMed] [Google Scholar]
  26. Logothetis NK, Pfeuffer J ( 2004): On the nature of the BOLD fMRI contrast mechanism. Magn Reson Imaging 22: 1517–1531. [DOI] [PubMed] [Google Scholar]
  27. Lowen SB, Nickerson LD, Levin JM ( 2009): Differential effects of acute cocaine and placebo administration on visual cortical activation in healthy subjects measured using BOLD fMRI. Pharmacol Biochem Behav 92: 277–282. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Macey PM, Macey KE, Kumar R, Harper RM ( 2004): A method for removal of global effects from fMRI time series. NeuroImage 22: 360–366. [DOI] [PubMed] [Google Scholar]
  29. Maquet P, Hirsch E, Metz‐Lutz MN, Motte J, Dive D, Marescaux C, Franck G ( 1995): Regional cerebral glucose metabolism in children with deterioration of one or more cognitive functions and continuous spike‐and‐wave discharges during sleep. Brain 118: 1497–1520. [DOI] [PubMed] [Google Scholar]
  30. Mbwana J, Berl MM, Ritzl EK, Rosenberger L, Mayo J, Weinstein S, Conry JA, Pearl PL, Shamim S, Moore EN, Sato S, Vezina LG, Theodore WH, Gaillard WD ( 2009): Limitations to plasticity of language network reorganization in localization related epilepsy. Brain 132: 347–356. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Möddel G, Lineweaver T, Schuele SU, Reinholz J, Loddenkemper T ( 2009): Atypical language lateralization in epilepsy patients. Epilepsia 50: 1505–1516. [DOI] [PubMed] [Google Scholar]
  32. Moeller F, Siebner HR, Ahlgrimm N, Wolff S, Muhle H, Granert O, Boor R, Jansen O, Gotman J, Stephani U, Siniatchkin M ( 2009): fMRI activation during spike and wave discharges evoked by photic stimulation. NeuroImage 48: 682–695. [DOI] [PubMed] [Google Scholar]
  33. Mühlau M, Wohlschläger AM, Gaser C, Valet M, Weindl A, Nunnemann S, Peinemann A, Etgen T, Ilg R ( 2009): Voxel‐based morphometry in individual patients: A pilot study in early Huntington disease. AJNR Am J Neuroradiol 30: 539–543. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Nakao Y, Itoh Y, Kuang TY, Cook M, Jehle J, Sokoloff L ( 2001): Effects of anesthesia on functional activation of cerebral blood flow and metabolism. Proc Natl Acad Sci USA 98: 7593–7598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Ogawa S, Lee TM ( 1990): Magnetic resonance imaging of blood vessels at high fields: In vivo and in vitro measurements and image simulation. Magn Reson Med 16: 9–18. [DOI] [PubMed] [Google Scholar]
  36. Oldfield RC ( 1971): The assessment and analysis of handedness: The Edinburgh inventory. Neuropsychologia 9: 97–113. [DOI] [PubMed] [Google Scholar]
  37. O'Shaughnessy ES, Berl MM, Moore EN, Gaillard WD ( 2008): Pediatric functional magnetic resonance imaging (fMRI): Issues and applications. J Child Neurol 23: 791–801. [DOI] [PubMed] [Google Scholar]
  38. Papanicolaou AC, Simos PG, Castillo EM, Breier JI, Sarkari S, Pataraia E, Billingsley RL, Buchanan S, Wheless J, Maggio V, Maggio WW ( 2004): Magnetoencephalography: A noninvasive alternative to the Wada procedure. J Neurosurg 100: 867–876. [DOI] [PubMed] [Google Scholar]
  39. Rasmussen T, Milner B ( 1977): The role of early left‐brain injury in determining lateralization of cerebral speech functions. Ann NY Acad Sci 299: 355–369. [DOI] [PubMed] [Google Scholar]
  40. Ricci GB, De Carli D, Colonnese C, Di Gennaro G, Quarato PP, Cantore G, Esposito V, Garreffa G, Maraviglia B ( 2004): Hemodynamic response (BOLD/fMRI) in focal epilepsy with reference to benzodiazepine effect. Magn Reson Imaging 22: 1487–1492. [DOI] [PubMed] [Google Scholar]
  41. Rosenberger LR, Zeck J, Berl MM, Moore EN, Ritzl EK, Shamim S, Weinstein SL, Conry JA, Pearl PL, Sato S, Vezina LG, Theodore WH, Gaillard WD ( 2009): Interhemispheric and intrahemispheric language reorganization in complex partial epilepsy. Neurology 72: 1830–1836. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ruff IM, Petrovich Brennan NM, Peck KK, Hou BL, Tabar V, Brennan CW, Holodny AI ( 2008): Assessment of the language laterality index in patients with brain tumor using functional MR imaging: Effects of thresholding, task selection, and prior surgery. AJNR Am J Neuroradiol 29: 528–535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Schapiro MB, Schmithorst VJ, Wilke M, Byars AW, Strawsburg RH, Holland SK ( 2004): BOLD fMRI signal increases with age in selected brain regions in children. Neuroreport 15: 2575–2578. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Staudt M, Grodd W, Niemann G, Wildgruber D, Erb M, Krägeloh‐Mann I ( 2001): Early left periventricular brain lesions induce right hemispheric organization of speech. Neurology 57: 122–125. [DOI] [PubMed] [Google Scholar]
  45. Stippich C, Rapps N, Dreyhaupt J, Durst A, Kress B, Nennig E, Tronnier VM, Sartor K ( 2007): Localizing and lateralizing language in patients with brain tumors: Feasibility of routine preoperative functional MR imaging in 81 consecutive patients. Radiology 243: 828–836. [DOI] [PubMed] [Google Scholar]
  46. Szaflarski JP, Binder JR, Possing ET, McKiernan KA, Ward BD, Hammeke TA ( 2002): Language lateralization in left‐handed and ambidextrous people: fMRI data. Neurology 59: 238–244. [DOI] [PubMed] [Google Scholar]
  47. Thulborn KR, Davis D, Erb P, Strojwas M, Sweeney JA ( 1996): Clinical fMRI: Implementation and experience. NeuroImage 4: S101–S107. [DOI] [PubMed] [Google Scholar]
  48. Van Halen L ( 2001): The Statistics of Variation In: Hall BK, editor. Variation. Amsterdam: Elsevier; pp 29–49. [Google Scholar]
  49. Voets NL, Adcock JE, Flitney DE, Behrens TE, Hart Y, Stacey R, Carpenter K, Matthews PM ( 2006): Distinct right frontal lobe activation in language processing following left hemisphere injury. Brain 129: 754–766. [DOI] [PubMed] [Google Scholar]
  50. Widdess‐Walsh P, Diehl B, Najm I ( 2006): Neuroimaging of focal cortical dysplasia. J Neuroimag 16: 185–196. [DOI] [PubMed] [Google Scholar]
  51. Wilke M, Schmithorst VJ ( 2006): A combined bootstrap/histogram analysis approach for computing a lateralization index from neuroimaging data. NeuroImage 33: 522–530. [DOI] [PubMed] [Google Scholar]
  52. Wilke M, Lidzba K ( 2007): LI‐tool: A new toolbox to assess lateralization in functional MR‐data. J Neurosci Methods 163: 128–136. [DOI] [PubMed] [Google Scholar]
  53. Wilke M, Schmithorst VJ, Holland SK ( 2002): Assessment of spatial normalization of whole‐brain magnetic resonance images in children. Hum Brain Mapp 17: 48–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Wilke M, Lidzba K, Staudt M, Buchenau K, Grodd W, Krägeloh‐Mann I ( 2005): Comprehensive language mapping in children, using fMRI: What's missing, counts. NeuroReport 16: 915–919. [DOI] [PubMed] [Google Scholar]
  55. Wilke M, Lidzba K, Staudt M, Buchenau K, Grodd W, Krägeloh‐Mann I ( 2006): An fMRI task battery for assessing hemispheric language dominance in children. NeuroImage 32: 400–410. [DOI] [PubMed] [Google Scholar]
  56. Wilke M, Holland SK, Altaye M, Gaser C ( 2008): Template‐O‐Matic: A toolbox for creating customized pediatric templates. NeuroImage 41: 903–913. [DOI] [PubMed] [Google Scholar]
  57. Wilke M, Staudt M, Juenger H, Grodd W, Braun C, Krägeloh‐Mann I ( 2009): Somatosensory system in two types of motor reorganization in congenital hemiparesis: Topography and function. Hum Brain Mapp 30: 776–788. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Wink AM, Roerdink JB ( 2004): Denoising functional MR images: A comparison of wavelet denoising and Gaussian smoothing. IEEE Trans Med Imaging 23: 374–387. [DOI] [PubMed] [Google Scholar]
  59. Yerys BE, Jankowski KF, Shook D, Rosenberger LR, Barnes KA, Berl MM, Ritzl EK, Vanmeter J, Vaidya CJ, Gaillard WD ( 2009): The fMRI success rate of children and adolescents: Typical development, epilepsy, attention deficit/hyperactivity disorder, and autism spectrum disorders. Hum Brain Mapp 30: 3426–3435. [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Human Brain Mapping are provided here courtesy of Wiley

RESOURCES