White-matter microstructure and language lateralization in left-handers: A whole-brain MRI analysis

Brain and Cognition 82 (2013) 319–328

Contents lists available at SciVerse ScienceDirect

Brain and Cognition

journal homepage: www.elsevier .com/ locate /b&c

White-matter microstructure and language lateralizationin left-handers: A whole-brain MRI analysis

0278-2626/$ - see front matter � 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.bandc.2013.05.005

⇑ Corresponding author. Fax: +36 72 535911.E-mail address: [email protected] (R. Horvath).

1 These authors contributed equally to this work.

Gabor Perlaki a,b,c,1, Reka Horvath a,⇑,1, Gergely Orsi a,b,c, Mihaly Aradi b, Tibor Auer d,e, Eszter Varga f,Gyongyi Kantor g, Anna Altbäcker a, Flora John a, Tamas Doczi c,h, Samuel Komoly a, Norbert Kovacs a,c,Attila Schwarcz c,h, Jozsef Janszky a,c

a Department of Neurology, University of Pecs, Ret u. 2, 7623 Pecs, Hungaryb Pecs Diagnostic Centre, Ret u. 2, 7623 Pecs, Hungaryc MTA-PTE Clinical Neuroscience MR Research Group, Ret u. 2, 7623 Pecs, Hungaryd Biomedizinische NMR Forschungs GmbH am Max-Planck Institut für Biophysikalische Chemie, Am Fassberg 11, 37070 Göttingen, Germanye Central Department of Radiology, Hospital Markusovszky, Markusovszky u. 5, 9700 Szombathely, Hungaryf Department of Psychiatry and Psychotherapy, University of Pecs, Ret u. 2, 7623 Pecs, Hungaryg Department of General Linguistics, University of Pecs, Ifjusag u. 6, 7624 Pecs, Hungaryh Department of Neurosurgery, University of Pecs, Ret u. 2, 7623 Pecs, Hungary

a r t i c l e i n f o a b s t r a c t

Article history:Accepted 20 May 2013Available online 21 June 2013

Keywords:Diffusion tensor imagingFunctional magnetic resonance imagingAtypical speech lateralityLeft-handersSuperior longitudinal fasciculusSuperior parietal lobe

Most people are left-hemisphere dominant for language. However the neuroanatomy of languagelateralization is not fully understood. By combining functional magnetic resonance imaging (fMRI) anddiffusion tensor imaging (DTI), we studied whether language lateralization is associated with cerebralwhite-matter (WM) microstructure. Sixteen healthy, left-handed women aged 20–25 were included inthe study. Left-handers were targeted in order to increase the chances of involving subjects with atypicallanguage lateralization. Language lateralization was determined by fMRI using a verbal fluency paradigm.Tract-based spatial statistics analysis of DTI data was applied to test for WM microstructural correlates oflanguage lateralization across the whole brain. Fractional anisotropy and mean diffusivity were used asindicators of WM microstructural organization. Right-hemispheric language dominance was associatedwith reduced microstructural integrity of the left superior longitudinal fasciculus and left-sided parietallobe WM. In left-handed women, reduced integrity of the left-sided language related tracts may be closelylinked to the development of right hemispheric language dominance. Our results may offer new insights intolanguage lateralization and structure–function relationships in human language system.

� 2013 Elsevier Inc. All rights reserved.

1. Introduction

The incidence of left-handedness, ranging from moderatethrough strong left-handed is about 10% in the population (Gilbert& Wysocki, 1992; Hardyck & Petrinovich, 1977). Several type offactors may affect the development of left-handedness includingmaternal handedness and familiar history of sinistrality (Annett,1983; Knecht et al., 2000; McKeever, 2000), gender (Gilbert & Wy-socki, 1992), testosterone level (Tan, 1991) and history of earlybrain injury (Satz, Orsini, Saslow, & Henry, 1985). Genetic theoriesof handedness and brain asymmetry were also suggested assuminga genetic predisposition toward right-handedness and left-hemi-spheric dominance for language (Annett, 2002; McManus, 2002).Left-handedness may be related to certain psychiatric diseases

(especially schizophrenia and bipolar disorders) (Klar, 1999; Satz& Green, 1999), developmental disorders (Dane & Balci, 2007; Goez& Zelnik, 2008) and central nervous system disorders (Llaurens,Raymond, & Faurie, 2009).

Left-hemispheric preference for language processing is one ofthe most consistent early findings (Auer et al., 2009; Josse & Tzou-rio-Mazoyer, 2004). Hemispheric language lateralization hasshown to be systematically associated with handedness: 94–96%of right-handed subjects showed left-hemispheric dominance,while 4–6% showed a bilateral pattern (Pujol, Deus, Losilla, & Cap-devila, 1999; Springer et al., 1999). In contrast, the incidence ofatypical (bilateral or right-hemispheric) language lateralizationwas found to increase in left-handers to approximately 15–30%(Josse & Tzourio-Mazoyer, 2004; Knecht et al., 2000; Pujol et al.,1999). Despite the fact that hemispheric language lateralizationof the human brain is a hot topic in neuroscience for a long time,its neuroanatomy is not fully understood (Propper et al., 2010;Vernooij et al., 2007).

320 G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328

Diffusion tensor imaging (DTI) is a technique in which at leastseven images must be acquired (six diffusion-weighted imagesalong non-collinear directions and one image with no diffusionweighting). From these images mean diffusivity (MD) and frac-tional anisotropy (FA) can be calculated, which represent the mag-nitude and directionality of water diffusion, respectively andprovide unique insights into white-matter (WM) microstructure.Increased MD may result from loss of barriers such as myelinsheaths, cell membranes or axons (Beaulieu, 2002), while FAthought to be related to tract integrity and may reflect the align-ment of neuronal fibers (Johansen-Berg & Behrens, 2009). HigherMD and lower FA values may indicate lower microstructuralorganization.

Some studies reported relationship between functional lan-guage lateralization and the morphometric magnetic resonanceimaging (MRI) findings of arcuate fasciculus (Propper et al.,2010), corpus callosum (Josse, Seghier, Kherif, & Price, 2008; Mof-fat, Hampson, & Lee, 1998), insula (Keller et al., 2011), right hippo-campus (Jansen et al., 2010) or left planum temporale (Josse,Mazoyer, Crivello, & Tzourio-Mazoyer, 2003; Moffat et al., 1998;Tzourio, Nkanga-Ngila, & Mazoyer, 1998). WM microstructure inrelation to language lateralization has also been recently investi-gated using the combination of functional magnetic resonanceimaging (fMRI) and DTI techniques (Häberling, Badzakova-Trajkov,& Corballis, 2011; Powell et al., 2006; Vernooij et al., 2007; West-erhausen et al., 2006). These studies tested for associations be-tween the lateralization of fMRI activations and various diffusiontensor measures (fractional anisotropy, relative anisotropy, rela-tive fiber density, mean diffusivity) in prespecified tracts of interesti.e. the corpus callosum and the arcuate fasciculus. Häberling et al.(2011) found that atypical hemispheric dominance for languagewas associated with higher mean fractional anisotropy (FA) ofthe corpus callosum, but this finding is somewhat in conflict withthose of Westerhausen et al. (2006), who found a trend towardshigher FA values for subjects with strongly left-lateralized lan-guage representation (Häberling et al., 2011; Westerhausen et al.,2006). In right-handed subjects both the relative fiber densityasymmetry index of arcuate fasciculus and the mean FA asymme-try index of connections between Broca’s and Wernicke’s areaswere found to be positively correlated with the degree of func-tional language dominance (Powell et al., 2006; Vernooij et al.,2007). A recent paper by Häberling et al. investigated the arcuatefasciculus asymmetry and its relationship to language dominancein monozygotic twins. They found that twin pairs with discordantlanguage dominance showed reversed asymmetry of anisotropicdiffusion in the arcuate fasciculus. The left-cerebrally dominanttwins showed leftward and the right-cerebrally dominant co-twinsshowed rightward asymmetry of anisotropic diffusion, indicating astrong nongenetic influence in arcuate fasciculus asymmetry(Häberling, Badzakova-Trajkov, & Corballis, 2013).

However, none of these studies provided insights into the rela-tionship of WM microstructure and language lateralization at thewhole-brain level. Tract-based spatial statistics (TBSS) is a new ap-proach aiming to improve the sensitivity, objectivity and interpret-ability of whole-brain analysis of multisubject diffusion data(Smith et al., 2006). TBSS attempts to combine the strengths ofvoxel-based morphometry-style (VBM-style) analysis (being ableto analyze the whole brain without prespecifying regions-of-inter-est) with the strengths of tractography-based methods (being con-fident that the estimates of FA are truly taken from the relevantvoxels). Unlike conventional VBM-style analysis, TBSS does not relystrongly on perfect cross-subject alignment or smoothing, thusallowing for unbiased whole-brain analysis of diffusion tensorproperties between multiple subjects.

In the present study, we investigated WM microstructure usingDTI at the whole-brain level (without prespecifying regions of

interest), in relation to hemispheric language lateralizationestablished with fMRI.

In order to increase the chances of including individuals exhib-iting atypical language lateralization (Szaflarski et al., 2002) and toavoid confounding effects of handedness (Buchel et al., 2004;Westerhausen et al., 2003, 2004), aging (Abe et al., 2008; Camara,Bodammer, Rodriguez-Fornells, & Tempelmann, 2007; Hsu et al.,2008) or sex-specific differences in WM integrity (Catani et al.,2007; Hagmann et al., 2006; Menzler et al., 2011; Thiebaut deSchotten et al., 2011) as well as the impact of interaction betweengender and handedness (Hagmann et al., 2006), only left-handed,healthy, young women were recruited. Our sample was as homog-enous as possible.

2. Methods

2.1. Subjects

Eighteen healthy, left-handed, Caucasian, female, graduate orpostgraduate university students between age of 20 and 25 with-out history of brain disorders were recruited through advertise-ments placed on notice boards across the University of Pécs. Thehandedness of all subjects was assessed by the Edinburgh Handed-ness Inventory (EHI) (Oldfield, 1971). Two individuals were ex-cluded due to excessive head movements during the fMRI and/orDTI protocols. Thus, we included a total of 16 healthy, young (meanage: 21.8 ± 1.7; range: 20–25 years), left-handed (mean EHI score:�79 ± 18.4; range: �50 to �100) women. The study was approvedby the Regional Ethical Committee of the University of Pécs andperformed in accordance with the ethical standards described inthe Declaration of Helsinki (Rickham, 1964). All subjects got de-tailed information about the investigation in both oral and writtenforms and gave written informed consent prior to the study.

2.2. Imaging data acquisition and visual analysis

All measurements were performed on a 3T Magnetom TIM Triohuman whole-body MRI scanner (Siemens AG, Erlangen, Germany)with a 12-channel head coil.

Functional images were acquired using a 2D single-shot gradi-ent-echo echo-planar imaging (EPI) sequence (TR/TE = 2000/36 ms; Flip Angle = 76�; 23 axial slices; slice thickness = 4 mm;no interslice gap; FOV = 192 � 192 mm2; matrix size = 92 � 92; re-ceiver bandwidth = 1360 Hz/pixel; interleaved slice order to avoidcrosstalk between contiguous slices). A total of 210 volumes wereacquired during the verbal fluency task.

DTI data were measured using a 2D single-shot diffusion-weighted spin-echo EPI sequence (TR/TE = 6700/78 ms; 60 axialslices; slice thickness = 2 mm; no interslice gap; FOV =211 � 260 mm2 matrix size = 104 � 128; diffusion gradients wereapplied in 20 directions with a b-value of 700 s/mm2 and a singlevolume was collected with no diffusion gradients applied; band-width = 1698 Hz/pixel; number of averages = 3).

Anatomical images were obtained using a T1-weighted three-dimensional MPRAGE sequence (TR/TI/TE = 1900/900/3.41 ms; FlipAngle = 9�; 144 axial slices; slice thickness = 0.9 mm; no interslicegap; FOV = 201 � 230 mm2; matrix size = 215 � 256; receiverbandwidth = 180 Hz/pixel).

Visual analysis of the images identified no brain abnormalities.

2.3. Functional MRI stimulation paradigm

As in our previous studies, a standard verbal fluency task with ablock design was used to assess functional hemispheric languagelateralization (Auer et al., 2009; Janszky et al., 2003). Although

G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328 321

there are some questions regarding the reliability of fMRI-basedlaterality assessment (Jansen et al., 2006), classification of lan-guage lateralization based on this task is consistent with the Wadatest, which is the ‘‘gold-standard’’ for measuring hemispheric lan-guage dominance (Adcock, Wise, Oxbury, Oxbury, & Matthews,2003; Woermann et al., 2003; Yetkin et al., 1998). Moreover, thefMRI asymmetry indices are reproducible in patients as well as incontrols (Adcock et al., 2003).

The paradigm included seven cycles of 30 second long restalternating with 30 second long internal word generation task(r-A-r-A-r-A-r-A-r-A-r-A-r-A; where r = rest and A = active phases).During the active conditions, the subjects were asked to silentlygenerate different words in Hungarian starting with a particularletter without any movements (e.g. pronunciation) until the word‘‘end’’ was announced indicating the rest phase. During the restperiods, they were instructed to stop the active task and relax.The seven different starting letters (S, K, E, T, L, A, N) were pre-sented via MRI-compatible electrostatic headphones specificallydesigned for fMRI (NordicNeuroLab, Bergen, Norway). Subjects laidin the scanner quietly with their eyes closed during both condi-tions. The whole fMRI examination was explained in detail beforethe scanning and subjects were questioned to ensure the instruc-tions were fully understood.

2.4. Functional MRI data processing and analysis

Pre-processing and statistical analysis were performed usingFEAT (FMRI Expert Analysis Tool) Version 5.98, part of FSL (FMRIB’sSoftware Library, www.fmrib.ox.ac.uk/fsl). Pre-processing includedBET brain extraction (Smith, 2002), MCFLIRT motion correction(Jenkinson, Bannister, Brady, & Smith, 2002), spatial smoothingwith 5 mm full width at half maximum and high-pass temporalfiltering with 120 s cut-off. The temporal filtering applied to thedata was used for the model as well. General linear model (GLM)time-series statistical analyses were carried out using FILM(FMRIB’s Improved Linear Model) with local autocorrelationcorrection (Woolrich, Ripley, Brady, & Smith, 2001). To model theblood oxygen level-dependent (BOLD) response a box-car functionwith ‘‘task’’ vs. ‘‘rest’’ conditions was convolved with the FSL’scanonical gamma haemodynamic response function (HRF). Thetemporal derivative of this waveform was also included in ourdesign-matrix to correct for slight overall temporal shifts betweenthe model and the data. First-level statistical maps were threshol-ded using clusters determined by Z > 2.3 and a corrected clustersignificance threshold of P = 0.05 to assess fMRI activations in mainlanguage areas (Worsley, 2001).

Single-session data sets were registered into standard spaceprior to language asymmetry index (AI) calculation using a three-step process. First, low-resolution fMRI data from each subjectwere registered to that subject’s (brain-extracted) high-resolutionstructural MRI (7 degrees-of-freedom linear fit), then the high-res-olution image was registered to the MNI152 standard brain image(12 degrees-of-freedom linear fit) using FLIRT (Jenkinson et al.,2002). Subsequently, the registration from structural image tothe standard space was further refined using FNIRT nonlinear reg-istration (Andersson, Jenkinson, & Smith, 2007a,b). The resultinglinear and non-linear deformations were combined mathemati-cally and applied to the first-level statistical maps to transformthem into standard space. Interpolation was only used in the finalstep, not in the registration calculations. Second-level mixed-ef-fects analyses (FLAME 1) were carried out to obtain the meangroup activations during verbal fluency task, separately for sub-jects with typical (n = 10) and atypical (n = 6) language lateraliza-tion. The resulting Z-score images were thresholded usingclusters determined by Z > 3.1 and a (corrected) cluster signifi-cance threshold of P = 0.01 for the typical and Z > 2.6 and a (cor-

rected) cluster significance threshold of P = 0.01 for the atypicalgroup (Worsley, 2001).

2.5. Language asymmetry index (AI) calculation

Because most areas of the frontal cortex are activated duringverbal fluency task (Auer et al., 2009), language lateralization anal-ysis was focused on frontal lobe. Anatomical region-of-interest(ROI) was generated in standard space for AI calculation by com-bining the frontal lobe defined by the 25% probability thresholdedMNI Structural Atlas with its mirrored image. The ROI was sym-metric across the cerebral hemispheres with respect to size, shapeand number of voxels to allow unbiased calculation of AI.

AIs were calculated using the LI-toolbox (Wilke & Lidzba, 2007)available as part of the SPM8 software package (Wellcome TrustCentre for Neuroimaging, London, UK, http://www.fil.ion.ucl.a-c.uk/spm/software/spm8/) for the frontal lobe mask created above(excluding the midline ±5 mm), employing the T-score mapsresulting from individual first-level analyses registered to standardspace as input (see Section 2.4). Because asymmetry indices basedon one single fixed statistical threshold do not yield robust orreproducible results (Jansen et al., 2006) and the asymmetry valueshighly depend on the applied threshold (Suarez, Whalen, O’Shea, &Golby, 2008), the robust and stable weighted bootstrapping meth-od was used to generate threshold-free AI values (Wilke & Schmi-thorst, 2006). Positive AI values represent left-hemisphericlanguage lateralization; negative ones indicate right-hemisphericlateralization.

Language dominance was classified as left hemispheric(AI > 0.2), bilateral (�0.2 6 AI 6 0.2) or right-sided (AI < �0.2)(Springer et al., 1999).

2.6. Voxelwise statistical analysis and diffusion data processing

Diffusion data were first corrected for eddy current distortionand head motion using a 12 parameter affine registration to a ref-erence volume (i.e. the volume without diffusion-weighting; b-va-lue = 0 s/mm2). After brain extraction of the diffusion data, FMRIB’sdiffusion toolbox (http://www.fmrib.ox.ac.uk/fsl/fdt) was used togenerate voxelwise images of fractional anisotropy (FA) and meandiffusivity (MD) by fitting a diffusion tensor model to the data ateach brain voxel (Smith et al., 2004).

Image analysis was carried out using Tract-Based Spatial Statis-tics (TBSS v1.2) (Smith et al., 2006). First, all subjects’ FA data werenonlinearly aligned to a high-resolution standard space average of58 well-aligned, good-quality FA images from healthy subjects(FMRIB58_FA standard space image) using the nonlinear registra-tion tool FNIRT (Andersson et al., 2007a,b). Next, the mean FA im-age was created and thinned to generate a mean FA skeleton whichrepresents the centers of all tracts common in all examined sub-jects. This mean skeleton was thresholded at FA > 0.3 to includemajor WM pathways where we could assume a good tract corre-spondence across subjects, but exclude minor tracts showing sub-stantial inter-subject variability and/or being potentiallydominated by grey matter or cerebrospinal fluid partial voluming(Smith et al., 2006). Individual FA images were then projected ontothis skeleton, by filling the skeleton with FA values from the near-est relevant tract center. This was achieved (for each skeleton vox-el) by searching perpendicular to the local skeleton structure forthe maximum value in the subject’s FA image. Using the samemean skeleton, a similar TBSS analysis was applied to the MD dataas well. The resulting skeletons were fed into voxelwise cross-sub-ject statistical analysis to test for linear correlations between waterdiffusion characteristics (FA, MD) and language lateralization.

To test the effect of language lateralization on asymmetry inwater diffusion, a symmetric mean FA skeleton was also derived

Table 1Asymmetry indices of the subjects.

Subject AI Language dominance

1 �0.65 Right-sided2 0.7 Left-sided3 �0.85 Right-sided4 0.41 Left-sided5 0.72 Left-sided6 �0.7 Right-sided7 �0.62 Right-sided8 �0.44 Right-sided9 0.33 Left-sided10 0.84 Left-sided11 0.79 Left-sided12 0.93 Left-sided13 �0.02 Bilateral14 0.84 Left-sided15 0.7 Left-sided16 0.9 Left-sided

AI = language asymmetry index evaluated from frontal lobe.

322 G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328

using ‘‘tbss_sym’’ script (http://www.fmrib.ox.ac.uk/fsl/tbss). Thisskeleton was restricted to only those parts of the ‘‘original’’ skele-ton that were already sufficiently close to being symmetric. Eachsubject’s aligned FA data were then projected onto this symmetricskeleton. To make the statistical test more straightforward, this 4Ddataset was left–right flipped, the latter subtracted from the for-mer and the difference was divided by the sum of the originaland flipped datasets. Finally, the left half of this dataset (corre-sponding to the ‘‘right’’ side of standard space) was zeroed, asthe same information (inverted) was present on the left and rightsides of the image. Asymmetry maps based on MD values were alsogenerated.

Voxelwise GLM was applied on the skeletonized diffusion datausing permutation-based non-parametric testing (5000 permuta-

Fig. 1. Group level activations during verbal fluency task separately for subjects with typusing clusters determined by Z > 3.1 and a (corrected) cluster significance threshold of PP = 0.01 for the atypical group. Axial slices are shown in radiological convention for MN

tions), with AI as variable of interest (Nichols & Holmes, 2002). Re-sults were considered significant for P < 0.05, corrected formultiple comparisons using ‘‘threshold-free cluster enhancement’’(TFCE), which avoids making an arbitrary choice of the cluster-forming threshold, while preserving the sensitivity benefits ofclusterwise correction (Smith & Nichols, 2009).

White-matter atlases provided with FSL version 4.1.8 (JHUICBM-DTI-81 white-matter labels atlas and JHU white-matter trac-tography atlas) and the second edition of the MRI Atlas of HumanWhite Matter (Oishi, Faria, Zijl, & Mori, 2010) were used to ana-tomically label the significant correlations in MNI space. Subse-quently, TBSS results were transformed back (back-projected)into the native space of subjects’ data and overlaid on the color-coded FA maps in order to provide further support for the atlas-based localizations and to confirm that the statistically significantskeleton points were derived from the same tract-center points inall subjects.

3. Results

3.1. Functional MRI

All 16 subjects included in the study performed successfully thesilent word generation task, as demonstrated by the robust activa-tion in the frontal language areas of at least one hemisphere. Lefthemispheric language lateralization was observed in 10 of 16 sub-jects, 1 showed bilateral and 5 demonstrated right-sided lateraliza-tion. Table 1 summarizes the asymmetry indices of theparticipants. Fig. 1a and Fig. 1b show the group level activationsduring internal word generation for subjects with typical (n = 10)and atypical (n = 6) language lateralization respectively. Anatomi-cal locations and coordinates of significant clusters are presentedin Table 2. In brief, significant activations were observed in the pre-central gyrus, inferior frontal gyrus, supplementary motor cortex,

ical (Fig. 1a) and atypical (Fig. 1b) language lateralization. Images were thresholded= 0.01 for the typical and Z > 2.6 and a (corrected) cluster significance threshold ofI slice coordinates from Z = �32 mm to Z = 60 mm.

Table 2Brain regions showing significant activations during internal word generation task insubjects with typical (a) and atypical (b) language lateralization.

ID Brain region Cluster size(mm3)

Z x y z

(a) Subjects with typical language lateralization1 Left precentral gyrus 26,344 5.46 �58 6 22

Left inferior frontal gyrus (p.opercularis)

4.82 �48 22 18

Left inferior frontal gyrus (p.triangularis)

4.53 �38 36 14

2 Left/right supplementarymotor cortex

6584 4.83 �4 �4 58

Left/right anterior cingulategyrus

3.84 �4 20 34

3 Left/right cerebellum 6408 4.45 �2 �58 �244 Right insular cortex 2296 4.19 34 16 2

(b) Subjects with atypical language lateralization1 Right insular cortex 9264 4.26 48 16 �6

Right inferior frontal gyrus(p. opercularis)

3.66 60 14 4

Right precentral gyrus 3.47 48 4 122 Left/right anterior cingulate

gyrus3976 3.39 12 18 36

Left/right supplementarymotor cortex

3.35 �8 �2 64

3 Left insular cortex 3200 3.77 �44 16 �64 Left/right cerebellum 3080 3.44 2 �56 �28

ID = cluster index; Z = Z-scores of local maxima in each cluster; x-, y- and z-valuescorrespond to the MNI coordinates of local maxima in mm; more local maxima arereported in each cluster when the cluster encompasses more than one anatomicallocation.

Fig. 2. Results of TBSS analysis of the WM skeleton. Red-yellow shows regions demasymmetry index (AI) and fractional anisotropy (FA). Blue-light blue indicates negativeregions are marked green. The skeletonized results have been thickened for better visibiliaxial slices), left AnG-WM (Z = 24–30 mm axial slices) and the left SPL-WM (Z = 42–50 mMatter (Oishi et al., 2010). Black shows mask of the mean FA skeleton thresholded at FA >slices are shown in radiological convention for MNI slice coordinates from Z = 16 mm toreader is referred to the web version of this article.)

G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328 323

anterior cingulate gyrus, cerebellum and insular cortex for bothgroups, but as expected activations were more strongly left-later-alized in subjects with AI > 0.2.

3.2. Voxelwise statistical analysis of diffusion data

TBSS analysis showed a significant inverse correlation betweenthe language asymmetry index (AI) and MD in the left superior lon-gitudinal fasciculus (SLF), left angular gyrus white-matter (AnG-WM) and the left superior parietal lobe white-matter (SPL-WM).Higher microstructural organization (lower MD) of these regionswas associated with left-hemispheric language dominance, whilelower structural organization (higher MD) was related to atypicallanguage lateralization (Fig. 2, Table 3a).

Local FA values also showed significant voxelwise correlationswith language lateralization in the left SLF and left AnG-WM. High-er microstructural organization (higher FA) was associated withleft-hemispheric language dominance, while lower structural orga-nization (lower FA) was related to atypical language lateralization(Fig. 2, Table 3b). A skeleton-based ROI analysis of FA was also per-formed in the left superior parietal region, where TBSS showed sig-nificant results only for MD (Fig. 2, axial slices from Z = 42 mmupwards). With the ROI analysis we could demonstrate a signifi-cant positive correlation between AI and FA for this region (Spear-man’s rank correlation, rho = 0.547, P = 0.028).

Scatterplots demonstrating the associations of FA vs. AI and MDvs. AI for the significant voxels found by TBSS are shown in Fig. 3aand b respectively.

onstrating significant positive correlation (corrected P < 0.05) between languagecorrelation (corrected P < 0.05) between AI and mean diffusivity (MD). Overlappingty. AI-related microstructural changes were identified in the left SLF (Z = 16–38 mmm axial slices) as labeled by the second edition of the MRI Atlas of Human White0.3. The background image is the group mean FA (in MNI152 standard space). AxialZ = 54 mm. (For interpretation of the references to color in this figure legend, the

Table 3Localization of significant correlations between language asymmetry index (AI) anddiffusion properties (MD and FA).

Brain region Cluster size(mm3)

P x y z

(a) Inverse correlation between AI and MDLeft superior longitudinal

fasciculus1088 0.024 �35 �36 33

Left superior parietal lobewhite-matter

0.027 �20 �48 41

Left angular gyrus white-matter 0.032 �35 �52 27

(b) Positive correlation between AI and FALeft superior longitudinal

fasciculus543 0.034 �36 �32 29

Left angular gyrus white-matter 0.037 �40 �46 23

P = TFCE-corrected significance levels of local maxima in each cluster; x-, y- and z-values correspond to the MNI coordinates of local maxima in mm; more localmaxima are reported in each cluster when the cluster encompasses more than oneanatomical location.

324 G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328

We did not find any correlation between language lateralizationand diffusion characteristics (FA, MD) in the corpus callosum,where earlier studies found contradictory results (see Section 1.).Furthermore, no linear correlations were seen between the lan-guage lateralization and brain asymmetry for FA or MD.

4. Discussion

The aim of the present study was to analyze the relationship be-tween functional hemispheric language lateralization and WMmicrostructure. To our knowledge this is the first study evaluatingthis relationship at the whole-brain level without any pre-specifi-cation of regions of interest using the unbiased voxel-based TBSSapproach. FA and MD – two quantitative indices of diffusion – wereused to assess WM microstructure. Water diffusion is hindered byseveral barriers such as myelin sheaths, cell membranes or axons(Beaulieu, 2002). Loss of structural barriers to water diffusion istypically accompanied by increased MD and decreased FA values.Higher MD and lower FA values may therefore indicate lowermicrostructural organization.

Fig. 3. Diffusion properties plotted against language asymmetry index (AI). Fig. 3a showsvoxels where TBSS showed significant association between AI and FA. Fig. 3b illustratesregion where TBSS showed significant negative relationship between AI and MD. Spearm

The main result of our study was that reduced microstructuralintegrity of the left-hemispheric WM was associated with atypicallanguage dominance.

4.1. Voxelwise statistical analysis of diffusion data

Our TBSS analysis showed that language lateralization was neg-atively correlated with MD and positively correlated with FA in theleft SLF. Subjects with stronger left-lateralized language functionhad lower diffusivity and higher anisotropy in the left SLF (morestructured SLF).

In terms of its volume, SLF may be the major cortical associationfiber pathway in the human brain running lateral to the corticospi-nal tract interconnecting frontal, temporal and parietal associationareas, which emphasizes its centrality to many associative or high-er order brain functions (Makris et al., 2005). Four different compo-nents of the SLF have been recently proposed (Makris et al., 2005).All of them have a frontal terminus in the posterior part of the fron-tal lobe, but they differ in their origin: SLF I originates in the supe-rior parietal lobe; SLF II and SLF III originate in the angular andsupramarginal gyri respectively. The forth subdivision of the SLF,the arcuate fasciculus stems from the caudal part of the superiortemporal gyrus arches around the caudal end of the Sylvian fissureand extends to the lateral prefrontal cortex along with the SLF II fi-bers. Although the arcuate fasciculus is just a component of thesuperior longitudinal fasciculus, their names are often inter-changed in the literature (Bernal & Altman, 2010; Bernal & Ardila,2009; Makris et al., 2005). According to the conventional model oflanguage organization, the arcuate fasciculus connects the lan-guage areas of Broca and Wernicke (Bernal & Altman, 2010; Catani& Mesulam, 2008). In spite of the abundant literature that accepts adirect connection between Broca’s and Wernicke’s areas (via thearcuate fasciculus) recent studies suggest a new language networkmodel emphasizing that the arcuate fasciculus connects the pos-terior brain areas with premotor/motor areas, and not directly withBroca’s area (Bernal & Altman, 2010; Bernal & Ardila, 2009). None-theless, the arcuate fasciculus may still connect to Broca’s areathrough a relay station located in the premotor/motor cortex. Aconnection like this could explain why the arcuate fasciculusseems to play a more important role in the speech (motor function)than in the language (cognition) (Bernal & Altman, 2010). Keeping

the positive correlation between AI and fractional anisotropy (FA) averaged over thethe negative correlation between AI and mean diffusivity (MD) averaged over thean’s correlation coefficients (rho) are also presented.

G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328 325

in mind that the arcuate fasciculus is likely to play a role in lan-guage, our result of association between language laterality andmicrostructural organization of the left SLF could be a reasonablefinding.

In consistent left-handers (after excluding of individuals withright-hemispheric language lateralization) Propper et al. found apositive correlation between the lateralization index of arcuate fas-ciculus volume and fMRI language lateralization in Wernicke’s area(Propper et al., 2010). Powell et al. demonstrated that right-handedsubjects with more highly lateralized mean FA of their fronto-tem-poral connections had more lateralized functional laterality of fMRIactivation for verb generation in the frontal lobe and for readingcomprehension in the temporal lobe (Powell et al., 2006). Signifi-cant correlation between the mean FA of the left-sided fronto-tem-poral connections and fMRI activity for reading comprehension inthe left inferior frontal gyrus was also reported, which may furthersupport our findings. Although the above cited articles may pro-vide modest support for our finding, these studies are not directlycomparable to our one.

Schizophrenic patients showed microstructural changes in theSLF (somewhat larger effect in the left SLF) compared to healthycontrols (Buchsbaum et al., 2006; Foong et al., 2000; Karlsgodtet al., 2008; Kong et al., 2011). Language lateralization was alsofound to be more atypical in schizophrenia (Sommer, Ramsey, &Kahn, 2001; van Veelen et al., 2011). Considering our present re-sults that microstructural organization of the left SLF are relatedto language lateralization in healthy subjects, we might hypothe-size that atypical language lateralization and altered WM micro-structure are not independent changes in schizophrenia.Moreover the current results indicate, that atypical language dom-inance and reduced WM integrity in the SLF may be a normalalthough rather rare deviation from the usual pattern which isnot necessarily associated with pathological states. However, lan-guage lateralization and WM microstructure has not been investi-gated simultaneously in schizophrenia. Future schizophreniastudies may benefit from conducting a combined analysis of DTIand language lateralization data and may resolve the cause/effectrelationship between structural and functional alterations.

The arcuate fasciculus belongs to the core perisylvian circuitryunderlying language, but other white matter tracts expandingthe boundaries of the canonical language network such as the infe-rior longitudinal fasciculus, the uncinate fasciculus and the inferiorfronto-occipital fasciculus may be relevant to language (Catani &Mesulam, 2008). However, our data indicate a lack of significanceof these additional regions in relation to language lateralization.

Language lateralization showed positive correlation with FAand negative correlation with MD in the left AnG-WM close tothe region recently termed as the ‘‘Geschwind’s territory’’ (Catani,Jones, & Ffytche, 2005). The Geschwind’s territory/inferior parietalcortex (including the angular gyrus) is suggested to have a mainrole in language comprehension and thought to be related tosemantic processing (Catani et al., 2005; Hart & Gordon, 1990;Price, 2000). Geschwind hypothesized that this area is involvedin the development of language because of its importance inenhancing cross-modal associations (Geschwind, 1965). Lesionsin this territory are associated with language related deficits, suchas alexia, agraphia (Price, 2000; Sakurai, Asami, & Mannen, 2010),dyslexia (Habib, 2000) or impairments on semantic comprehen-sion tasks (Hart & Gordon, 1990; Price, 2000). Functional magneticresonance imaging studies demonstrated greater angular gyrusBOLD-signal in response to the processing of words than non-words (Binder, Medler, Desai, Conant, & Liebenthal, 2005; Binderet al., 2003; Rissman, Eliassen, & Blumstein, 2003) and for normalcompared to semantically incongruent sentences (Humphries, Bin-der, Medler, & Liebenthal, 2007). Although, the relevance of angu-lar gyrus is obvious, the specific function of this area in language

comprehension remains to be clarified (Brownsett & Wise, 2010).It is usually considered as a part of a distributed language network,but it is also conceivable that this region is repeatedly observed incertain studies of language, because it is a part of a distributedworking memory system required for language comprehension(Brownsett & Wise, 2010). Although our present result alone can-not distinguish between these alternative models, it may imply aclose relationship between AnG-WM mictrostructure and languagelaterality.

Additionally an inverse correlation was found between lan-guage lateralization and MD in the left SPL-WM. Using whole-brain TBSS analysis the FA showed no correlation in this area,which seemed somewhat surprising taking into account thatMD and FA values – two most commonly used measures ofWM integrity – are generally inversely correlated. We conducteda skeleton-based ROI analysis to further investigate the superiorparietal region where MD showed significant correlation. Withthe ROI analysis we could demonstrate a significant positive asso-ciation between language lateralization and FA (Spearman’s rankcorrelation, rho = 0.547, P = 0.028). Although left superior parietallobe has an essential role in writing (Menon & Desmond, 2001), itis not a typical language area; it is rather related to sensorimotorintegration by maintaining an internal representation of the stateof both the world and one’s own body (Wolpert, Goodbody, &Husain, 1998). It has also been reported previously that parietallobe functional specialization is correlated with language lateral-ization, however it may neither explain this finding altogether(Badzakova-Trajkov, Häberling, Roberts, & Corballis, 2010). Inte-grating this result into current theories of language is not easy,but our observation in the left SPL-WM may suggest an importantdirection for future research of WM microstructural changesunderlying language lateralization. It is also possible that complexfiber crossings with other relevant (language-related) white-mat-ter fibers at this level, such as the SLF, could be responsible forthis correlation.

Conclusively, the voxel-based analysis showed a strongrelationship between left-hemispheric WM microstructure andlanguage lateralization in left-handed subjects. Subjects withleft-hemispheric language representation demonstrated morestructured left-sided WM than subjects with atypical languagelateralization. One possible reason for the relationship betweenleft-hemispheric WM and atypical language-lateralization mightlie in interhemispheric inhibition, suggesting that left-hemisphericinhibition of right hemisphere may partly contribute to languagedominance on the left side (Thiel et al., 2006). Reduced WM integ-rity of the left-hemispheric language related tracts may imply lessinhibition of right-sided language areas. We found no evidencethat right-hemispheric WM microstructure is related to languagelateralization, which could be a reflective of our restricted sample.While not excluding the possibility that right-sided WM integritymay be a marker of hemispheric specialization for language, weemphasize the importance of left-hemispheric WM integrity.

4.2. Methodological considerations

There was no behavioral measure of the fMRI task performanceduring image acquisition or alternative version of the task admin-istered outside the scanner. However, the applied fMRI task wasconsidered relatively easy and equally performable for healthyyoung university students, so disturbing effects of difference inperformance were not expected.

We claimed that our findings are free from potential effects ofaging, handedness and gender on WM microstructure and/orlanguage lateralization, thus only left-handed healthy youngwomen were enrolled. However, it is possible that our results onthis homogeneous group are not representative for the whole

326 G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328

population. We found that about 60% of our subjects (10 out of 16)showed left-hemisphere dominance for language, which issomewhat lower compared to other studies reporting 70–85% ofleft-handers with left-hemispheric language lateralization (Josse& Tzourio-Mazoyer, 2004; Knecht et al., 2000; Pujol et al., 1999).This difference may be related to our small sample size or couldbe a consequence of the fact that only women were included inthe present study (whereas previous data based on both womenand men). Some studies indicate that language is less lateralizedin women (Kansaku, Yamaura, & Kitazawa, 2000; Phillips, Lowe,Lurito, Dzemidzic, & Mathews, 2001; Vikingstad, George, Johnson,& Cao, 2000), whereas others found no sex difference in languagelateralization (Knecht et al., 2000; Sommer, Aleman, Bouma, &Kahn, 2004).

The ROI used for AI calculation was symmetric across thecerebral hemispheres. The aim was to allow unbiased calcula-tion of AI with the applied threshold-independent method (thatrequires equal number of voxels on both hemispheres),although we recognize the limitations of this approach giventhat frontal lobe is asymmetric in most individuals and thatstructurally homotopic regions do not necessarily correspondfunctionally.

A possible limitation of using DTI data is that the standard ten-sor model of DTI cannot adequately characterize complex fiberarchitectures (when an image voxel contains fiber populationswith more than one dominant orientation), therefore the direc-tionally dependent diffusion measure (FA) as well as the scalarmeasure of the total diffusion within a voxel (MD) may be influ-enced (Vos, Jones, Jeurissen, Viergever, & Leemans, 2012). This is-sue is not limited to our methodology only (Cubon, Putukian,Boyer, & Dettwiler, 2011; Giorgio et al., 2010; Li et al., 2010; Tak-ao, Hayashi, & Ohtomo, 2011) and is due to the fact that we areunable to noninvasively resolve fiber tracts at a microscopic level.Even though some methods have been developed to estimate anddifferentiate multiple fibers in each voxel (Behrens, Berg, Jbabdi,Rushworth, & Woolrich, 2007; Tuch, Reese, Wiegell, & Wedeen,2003), the challenge of complex fiber structures has not beencompletely resolved. It is likely that the centers of fiber bundlesanalysed by TBSS are more compactly packed with less complexfiber orientation, thus our results do not seem to be muchinfluenced.

MR markers of diffusion (e.g. FA or MD) have to be interpretedcarefully, because they are sensitive to several tissue characteris-tics and there is no one-to-one relationship between a given ana-tomical property and a particular MR measure (Johansen-Berg &Behrens, 2009).

5. Conclusions

In left-handed women, we found strong support that left SLFand left-sided parietal lobe WM microstructure are related to func-tional language lateralization. Based on our data, it may be hypoth-esized that reduced white-matter integrity in the left-sidedlanguage related tracts are closely linked to the development ofright hemispheric language dominance. The dependence of theright-sided language lateralization on the integrity of left-hemi-spheric pathways but not the right-hemispheric ones is not easilyexplained, but might be a useful reference for further investiga-tions on how atypical language dominance develops. Microstruc-tural organization of cerebral WM may be a predictor oflanguage laterality, but the cause/effect problem between alteredWM microstructure and language lateralization is not resolved inthis study. Our results may offer new insights into language later-alization and structure–function relationships in human languagesystem.

Disclosure statement

We are not aware of any actual or potential conflicts of interestregarding this work.

Acknowledgments

This work was supported by the Hungarian Research Fund(OTKA F68720; OTKA K105357), the Hungarian Research Council(ETT 272/2009), PTE AOK-KA-2013/12 and the grant ‘‘DevelopingCompetitiveness of Universities in the South Transdanubian Region(SROP-4.2.1.B-10/2/KONV-2010-0002)’’. JJ, AS and NK weresupported by the Bolyai Scholarship of the Hungarian Academyof Science. NK was also supported by the PTE AOK-KA No: 2011/34039/KA-OTKA/11-10. Financial support by the German FederalMinistry for Education and Research (BMBF) via the BernsteinFocus Neurotechnologie (BFNT) Göttingen (Grant no. 01GQ0812)is gratefully acknowledged.

References

Abe, O., Yamasue, H., Aoki, S., Suga, M., Yamada, H., Kasai, K., et al. (2008). Aging inthe CNS: Comparison of gray/white matter volume and diffusion tensor data.Neurobiology of Aging, 29(1), 102–116.

Adcock, J. E., Wise, R. G., Oxbury, J. M., Oxbury, S. M., & Matthews, P. M. (2003).Quantitative fMRI assessment of the differences in lateralization of language-related brain activation in patients with temporal lobe epilepsy. Neuroimage,18(2), 423–438.

Andersson, J. L. R., Jenkinson, M., & Smith, S. M. (2007a). Non-linear optimisation.FMRIB technical report TR07JA1.

Andersson, J. L. R., Jenkinson, M., & Smith, S. M. (2007b). Non-linear registration, akaspatial normalisation. FMRIB technical report TR07JA2.

Annett, M. (1983). Hand preference and skill in 115 children of two left-handedparents. British Journal of Psychology, 74(Pt 1), 17–32.

Annett, M. (2002). Handedness and brain asymmetry. Psychology Press.Auer, T., Pinter, S., Kovacs, N., Kalmar, Z., Nagy, F., Horvath, R. A., et al. (2009). Does

obstetric brachial plexus injury influence speech dominance? Annals ofNeurology, 65(1), 57–66.

Badzakova-Trajkov, G., Häberling, I. S., Roberts, R. P., & Corballis, M. C. (2010).Cerebral asymmetries: Complementary and independent processes. PLoS ONE,5(3), e9682.

Beaulieu, C. (2002). The basis of anisotropic water diffusion in the nervous system –A technical review. NMR in Biomedicine, 15(7–8), 435–455.

Behrens, T. E., Berg, H. J., Jbabdi, S., Rushworth, M. F., & Woolrich, M. W. (2007).Probabilistic diffusion tractography with multiple fibre orientations: What canwe gain? Neuroimage, 34(1), 144–155.

Bernal, B., & Altman, N. (2010). The connectivity of the superior longitudinalfasciculus: A tractography DTI study. Magnetic Resonance Imaging, 28(2),217–225.

Bernal, B., & Ardila, A. (2009). The role of the arcuate fasciculus in conductionaphasia. Brain, 132(Pt 9), 2309–2316.

Binder, J. R., McKiernan, K. A., Parsons, M. E., Westbury, C. F., Possing, E. T., Kaufman,J. N., et al. (2003). Neural correlates of lexical access during visual wordrecognition. Journal of Cognitive Neuroscience, 15(3), 372–393.

Binder, J. R., Medler, D. A., Desai, R., Conant, L. L., & Liebenthal, E. (2005). Someneurophysiological constraints on models of word naming. Neuroimage, 27(3),677–693.

Brownsett, S. L., & Wise, R. J. (2010). The contribution of the parietal lobes tospeaking and writing. Cerebral Cortex, 20(3), 517–523.

Buchel, C., Raedler, T., Sommer, M., Sach, M., Weiller, C., & Koch, M. A. (2004). Whitematter asymmetry in the human brain: A diffusion tensor MRI study. CerebralCortex, 14(9), 945–951.

Buchsbaum, M. S., Friedman, J., Buchsbaum, B. R., Chu, K. W., Hazlett, E. A.,Newmark, R., et al. (2006). Diffusion tensor imaging in schizophrenia. BiologicalPsychiatry, 60(11), 1181–1187.

Camara, E., Bodammer, N., Rodriguez-Fornells, A., & Tempelmann, C. (2007). Age-related water diffusion changes in human brain: A voxel-based approach.Neuroimage, 34(4), 1588–1599.

Catani, M., Allin, M. P., Husain, M., Pugliese, L., Mesulam, M. M., Murray, R. M., et al.(2007). Symmetries in human brain language pathways correlate with verbalrecall. Proceedings of the National Academy of Sciences of the United States ofAmerica, 104(43), 17163–17168.

Catani, M., Jones, D. K., & Ffytche, D. H. (2005). Perisylvian language networks of thehuman brain. Annals of Neurology, 57(1), 8–16.

Catani, M., & Mesulam, M. (2008). The arcuate fasciculus and the disconnectiontheme in language and aphasia: History and current state. Cortex, 44(8),953–961.

Cubon, V. A., Putukian, M., Boyer, C., & Dettwiler, A. (2011). A diffusion tensorimaging study on the white matter skeleton in individuals with sports-relatedconcussion. Journal of Neurotrauma, 28(2), 189–201.

G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328 327

Dane, S., & Balci, N. (2007). Handedness, eyedness and nasal cycle in children withautism. International Journal of Developmental Neuroscience, 25(4), 223–226.

Foong, J., Maier, M., Clark, C. A., Barker, G. J., Miller, D. H., & Ron, M. A. (2000).Neuropathological abnormalities of the corpus callosum in schizophrenia: Adiffusion tensor imaging study. Journal of Neurology Neurosurgery and Psychiatry,68(2), 242–244.

Geschwind, N. (1965). Disconnexion syndromes in animals and man. I. Brain, 88(2),237–294.

Gilbert, A. N., & Wysocki, C. J. (1992). Hand preference and age in the United States.Neuropsychologia, 30(7), 601–608.

Giorgio, A., Palace, J., Johansen-Berg, H., Smith, S. M., Ropele, S., Fuchs, S., et al.(2010). Relationships of brain white matter microstructure with clinical and MRmeasures in relapsing-remitting multiple sclerosis. Journal of MagneticResonance Imaging, 31(2), 309–316.

Goez, H., & Zelnik, N. (2008). Handedness in patients with developmentalcoordination disorder. Journal of Child Neurology, 23(2), 151–154.

Häberling, I. S., Badzakova-Trajkov, G., & Corballis, M. C. (2011). Callosal tracts andpatterns of hemispheric dominance: A combined fMRI and DTI study.Neuroimage, 54(2), 779–786.

Häberling, I. S., Badzakova-Trajkov, G., & Corballis, M. C. (2013). Asymmetries of thearcuate fasciculus in monozygotic twins: Genetic and nongenetic influences.PLoS ONE, 8(1), e52315.

Habib, M. (2000). The neurological basis of developmental dyslexia: An overviewand working hypothesis. Brain, 123(Pt 12), 2373–2399.

Hagmann, P., Cammoun, L., Martuzzi, R., Maeder, P., Clarke, S., Thiran, J. P., et al.(2006). Hand preference and sex shape the architecture of language networks.Human Brain Mapping, 27(10), 828–835.

Hardyck, C., & Petrinovich, L. F. (1977). Left-handedness. Psychological Bulletin,84(3), 385–404.

Hart, J., Jr., & Gordon, B. (1990). Delineation of single-word semantic comprehensiondeficits in aphasia, with anatomical correlation. Annals of Neurology, 27(3),226–231.

Hsu, J. L., Leemans, A., Bai, C. H., Lee, C. H., Tsai, Y. F., Chiu, H. C., et al. (2008). Genderdifferences and age-related white matter changes of the human brain: Adiffusion tensor imaging study. Neuroimage, 39(2), 566–577.

Humphries, C., Binder, J. R., Medler, D. A., & Liebenthal, E. (2007). Time course ofsemantic processes during sentence comprehension: An fMRI study.Neuroimage, 36(3), 924–932.

Jansen, A., Liuzzi, G., Deppe, M., Kanowski, M., Olschlager, C., Albers, J. M., et al.(2010). Structural correlates of functional language dominance: A voxel-basedmorphometry study. Journal of Neuroimaging, 20(2), 148–156.

Jansen, A., Menke, R., Sommer, J., Forster, A. F., Bruchmann, S., Hempleman, J., et al.(2006). The assessment of hemispheric lateralization in functional MRI-robustness and reproducibility. Neuroimage, 33(1), 204–217.

Janszky, J., Ebner, A., Kruse, B., Mertens, M., Jokeit, H., Seitz, R. J., et al. (2003).Functional organization of the brain with malformations of corticaldevelopment. Annals of Neurology, 53(6), 759–767.

Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization forthe robust and accurate linear registration and motion correction of brainimages. Neuroimage, 17(2), 825–841.

Johansen-Berg, H., & Behrens, T. E. J (2009). Diffusion MRI: Elsevier Science.Josse, G., Mazoyer, B., Crivello, F., & Tzourio-Mazoyer, N. (2003). Left planum

temporale: An anatomical marker of left hemispheric specialization forlanguage comprehension. Brain Research. Cognitive Brain Research, 18(1), 1–14.

Josse, G., Seghier, M. L., Kherif, F., & Price, C. J. (2008). Explaining function withanatomy: Language lateralization and corpus callosum size. Journal ofNeuroscience, 28(52), 14132–14139.

Josse, G., & Tzourio-Mazoyer, N. (2004). Hemispheric specialization for language.Brain Research. Brain Research Reviews, 44(1), 1–12.

Kansaku, K., Yamaura, A., & Kitazawa, S. (2000). Sex differences in lateralizationrevealed in the posterior language areas. Cerebral Cortex, 10(9), 866–872.

Karlsgodt, K. H., van Erp, T. G. M., Poldrack, R. A., Bearden, C. E., Nuechterlein, K. H., &Cannon, T. D. (2008). Diffusion tensor imaging of the superior longitudinalfasciculus and working memory in recent-onset schizophrenia. BiologicalPsychiatry, 63(5), 512–518.

Keller, S. S., Roberts, N., Garcia-Finana, M., Mohammadi, S., Ringelstein, E. B., Knecht,S., et al. (2011). Can the language-dominant hemisphere be predicted by brainanatomy? Journal of Cognitive Neuroscience, 23(8), 2013–2029.

Klar, A. J. (1999). Genetic models for handedness, brain lateralization,schizophrenia, and manic-depression. Schizophrenia Research, 39(3), 207–218.

Knecht, S., Drager, B., Deppe, M., Bobe, L., Lohmann, H., Floel, A., et al. (2000).Handedness and hemispheric language dominance in healthy humans. Brain,123(Pt 12), 2512–2518.

Kong, X. J., Ouyang, X. A., Tao, H. J., Liu, H. H., Li, L., Zhao, J. P., et al. (2011).Complementary diffusion tensor imaging study of the corpus callosum inpatients with first-episode and chronic schizophrenia. Journal of Psychiatry &Neuroscience, 36(2), 120–125.

Li, L., Preuss, T. M., Rilling, J. K., Hopkins, W. D., Glasser, M. F., Kumar, B., et al. (2010).Chimpanzee (Pan troglodytes) precentral corticospinal system asymmetry andhandedness: A diffusion magnetic resonance imaging study. PLoS ONE, 5(9),e12886.

Llaurens, V., Raymond, M., & Faurie, C. (2009). Why are some people left-handed?An evolutionary perspective. Philosophical Transactions of the Royal Society ofLondon. Series B, Biological sciences, 364(1519), 881–894.

Makris, N., Kennedy, D. N., McInerney, S., Sorensen, A. G., Wang, R., Caviness, V. S.Jr.,, et al. (2005). Segmentation of subcomponents within the superior

longitudinal fascicle in humans: A quantitative, in vivo, DT-MRI study.Cerebral Cortex, 15(6), 854–869.

McKeever, W. F. (2000). A new family handedness sample with findings consistentwith X-linked transmission. British Journal of Psychology, 91(Pt 1), 21–39.

McManus, I. C (2002). Right hand, left hand: The origins of asymmetry in brains, bodies,atoms, and cultures. Harvard University Press.

Menon, V., & Desmond, J. E. (2001). Left superior parietal cortex involvement inwriting: Integrating fMRI with lesion evidence. Brain Research. Cognitive BrainResearch, 12(2), 337–340.

Menzler, K., Belke, M., Wehrmann, E., Krakow, K., Lengler, U., Jansen, A., et al. (2011).Men and women are different: Diffusion tensor imaging reveals sexualdimorphism in the microstructure of the thalamus, corpus callosum andcingulum. Neuroimage, 54(4), 2557–2562.

Moffat, S. D., Hampson, E., & Lee, D. H. (1998). Morphology of the planum temporaleand corpus callosum in left handers with evidence of left and right hemispherespeech representation. Brain, 121(Pt 12), 2369–2379.

Nichols, T. E., & Holmes, A. P. (2002). Nonparametric permutation tests forfunctional neuroimaging: A primer with examples. Human Brain Mapping,15(1), 1–25.

Oishi, K., Faria, A. V., Zijl, P. C. M., & Mori, S. (2010). MRI atlas of human white matter.Elsevier Science.

Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburghinventory. Neuropsychologia, 9(1), 97–113.

Phillips, M. D., Lowe, M. J., Lurito, J. T., Dzemidzic, M., & Mathews, V. P. (2001).Temporal lobe activation demonstrates sex-based differences during passivelistening. Radiology, 220(1), 202–207.

Powell, H. W., Parker, G. J., Alexander, D. C., Symms, M. R., Boulby, P. A., Wheeler-Kingshott, C. A., et al. (2006). Hemispheric asymmetries in language-relatedpathways: A combined functional MRI and tractography study. Neuroimage,32(1), 388–399.

Price, C. J. (2000). The anatomy of language: Contributions from functionalneuroimaging. Journal of Anatomy, 197(Pt 3), 335–359.

Propper, R. E., O’Donnell, L. J., Whalen, S., Tie, Y., Norton, I. H., Suarez, R. O., et al.(2010). A combined fMRI and DTI examination of functional languagelateralization and arcuate fasciculus structure: Effects of degree versusdirection of hand preference. Brain and Cognition, 73(2), 85–92.

Pujol, J., Deus, J., Losilla, J. M., & Capdevila, A. (1999). Cerebral lateralization oflanguage in normal left-handed people studied by functional MRI. Neurology,52(5), 1038–1043.

Rickham, P. P. (1964). Human Experimentation. Code of Ethics of the World MedicalAssociation. Declaration of Helsinki.. Br Med J, 2(5402), 177.

Rissman, J., Eliassen, J. C., & Blumstein, S. E. (2003). An event-related FMRIinvestigation of implicit semantic priming. Journal of Cognitive Neuroscience,15(8), 1160–1175.

Sakurai, Y., Asami, M., & Mannen, T. (2010). Alexia and agraphia with lesions of theangular and supramarginal gyri: Evidence for the disruption of sequentialprocessing. Journal of the Neurological Sciences, 288(1–2), 25–33.

Satz, P., & Green, M. F. (1999). Atypical handedness in schizophrenia: Somemethodological and theoretical issues. Schizophrenia Bulletin, 25(1), 63–78.

Satz, P., Orsini, D. L., Saslow, E., & Henry, R. (1985). The pathological left-handednesssyndrome. Brain and Cognition, 4(1), 27–46.

Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping,17(3), 143–155.

Smith, S. M., Jenkinson, M., Johansen-Berg, H., Rueckert, D., Nichols, T. E., Mackay, C.E., et al. (2006). Tract-based spatial statistics: Voxelwise analysis of multi-subject diffusion data. Neuroimage, 31(4), 1487–1505.

Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E.,Johansen-Berg, H., et al. (2004). Advances in functional and structural MR imageanalysis and implementation as FSL. Neuroimage, 23(Suppl. 1), S208–219.

Smith, S. M., & Nichols, T. E. (2009). Threshold-free cluster enhancement:Addressing problems of smoothing, threshold dependence and localisation incluster inference. Neuroimage, 44(1), 83–98.

Sommer, I. E., Aleman, A., Bouma, A., & Kahn, R. S. (2004). Do women really havemore bilateral language representation than men? A meta-analysis offunctional imaging studies. Brain, 127(Pt 8), 1845–1852.

Sommer, I. E., Ramsey, N. F., & Kahn, R. S. (2001). Language lateralization inschizophrenia, an fMRI study. Schizophrenia Research, 52(1–2), 57–67.

Springer, J. A., Binder, J. R., Hammeke, T. A., Swanson, S. J., Frost, J. A., Bellgowan, P.S., et al. (1999). Language dominance in neurologically normal and epilepsysubjects: A functional MRI study. Brain, 122(Pt 11), 2033–2046.

Suarez, R. O., Whalen, S., O’Shea, J. P., & Golby, A. J. (2008). A surgical planningmethod for functional MRI assessment of language dominance: Influences fromthreshold, region-of-interest, and stimulus mode. Brain Imaging and Behavior,2(2), 59–73.

Szaflarski, J. P., Binder, J. R., Possing, E. T., McKiernan, K. A., Ward, B. D., & Hammeke,T. A. (2002). Language lateralization in left-handed and ambidextrous people:fMRI data. Neurology, 59(2), 238–244.

Takao, H., Hayashi, N., & Ohtomo, K. (2011). White matter asymmetry in healthyindividuals: A diffusion tensor imaging study using tract-based spatialstatistics. Neuroscience, 193, 291–299.

Tan, U. (1991). Serum testosterone levels in male and female subjects with standardand anomalous dominance. International Journal of Neuroscience, 58(3–4), 211–214.

Thiebaut de Schotten, M., Ffytche, D. H., Bizzi, A., Dell’Acqua, F., Allin, M., Walshe,M., et al. (2011). Atlasing location, asymmetry and inter-subject variability ofwhite matter tracts in the human brain with MR diffusion tractography.Neuroimage, 54(1), 49–59.

328 G. Perlaki et al. / Brain and Cognition 82 (2013) 319–328

Thiel, A., Schumacher, B., Wienhard, K., Gairing, S., Kracht, L. W., Wagner, R., et al.(2006). Direct demonstration of transcallosal disinhibition in languagenetworks. Journal of Cerebral Blood Flow and Metabolism, 26(9), 1122–1127.

Tuch, D. S., Reese, T. G., Wiegell, M. R., & Wedeen, V. J. (2003). Diffusion MRI ofcomplex neural architecture. Neuron, 40(5), 885–895.

Tzourio, N., Nkanga-Ngila, B., & Mazoyer, B. (1998). Left planum temporale surfacecorrelates with functional dominance during story listening. NeuroReport, 9(5),829–833.

van Veelen, N. M., Vink, M., Ramsey, N. F., Sommer, I. E., van Buuren, M.,Hoogendam, J. M., et al. (2011). Reduced language lateralization in first-episodemedication-naive schizophrenia. Schizophrenia Research, 127(1–3), 195–201.

Vernooij, M. W., Smits, M., Wielopolski, P. A., Houston, G. C., Krestin, G. P., & van derLugt, A. (2007). Fiber density asymmetry of the arcuate fasciculus in relation tofunctional hemispheric language lateralization in both right- and left-handedhealthy subjects: A combined fMRI and DTI study. Neuroimage, 35(3),1064–1076.

Vikingstad, E. M., George, K. P., Johnson, A. F., & Cao, Y. (2000). Cortical languagelateralization in right handed normal subjects using functional magneticresonance imaging. Journal of the Neurological Sciences, 175(1), 17–27.

Vos, S. B., Jones, D. K., Jeurissen, B., Viergever, M. A., & Leemans, A. (2012). Theinfluence of complex white matter architecture on the mean diffusivity indiffusion tensor MRI of the human brain. Neuroimage, 59(3), 2208–2216.

Westerhausen, R., Kreuder, F., Dos Santos Sequeira, S., Walter, C., Woerner, W.,Wittling, R. A., et al. (2004). Effects of handedness and gender on macro- andmicrostructure of the corpus callosum and its subregions: A combined high-resolution and diffusion-tensor MRI study. Brain Research. Cognitive BrainResearch, 21(3), 418–426.

Westerhausen, R., Kreuder, F., Dos Santos Sequeira, S., Walter, C., Woerner, W.,Wittling, R. A., et al. (2006). The association of macro- and microstructure ofthe corpus callosum and language lateralisation. Brain and Language, 97(1),80–90.

Westerhausen, R., Walter, C., Kreuder, F., Wittling, R. A., Schweiger, E., & Wittling,W. (2003). The influence of handedness and gender on the microstructure of thehuman corpus callosum: A diffusion-tensor magnetic resonance imaging study.Neuroscience Letters, 351(2), 99–102.

Wilke, M., & Lidzba, K. (2007). LI-tool: A new toolbox to assess lateralization infunctional MR-data. Journal of Neuroscience Methods, 163(1), 128–136.

Wilke, M., & Schmithorst, V. J. (2006). A combined bootstrap/histogram analysisapproach for computing a lateralization index from neuroimaging data.Neuroimage, 33(2), 522–530.

Woermann, F. G., Jokeit, H., Luerding, R., Freitag, H., Schulz, R., Guertler, S., et al.(2003). Language lateralization by Wada test and fMRI in 100 patients withepilepsy. Neurology, 61(5), 699–701.

Wolpert, D. M., Goodbody, S. J., & Husain, M. (1998). Maintaining internalrepresentations: The role of the human superior parietal lobe. NatureNeuroscience, 1(6), 529–533.

Woolrich, M. W., Ripley, B. D., Brady, M., & Smith, S. M. (2001). Temporalautocorrelation in univariate linear modeling of FMRI data. Neuroimage, 14(6),1370–1386.

Worsley, K. J. (2001). Statistical analysis of activation images. Functional MRI: AnIntroduction to Methods, pp. 251–270.

Yetkin, F. Z., Swanson, S., Fischer, M., Akansel, G., Morris, G., Mueller, W., et al.(1998). Functional MR of frontal lobe activation: Comparison with Wadalanguage results. AJNR. American Journal of Neuroradiology, 19(6), 1095–1098.