Not on speaking terms: hallucinations and structural network disconnectivity in schizophrenia

ORIGINAL ARTICLE

Not on speaking terms: hallucinations and structural networkdisconnectivity in schizophrenia

Branislava Curcic-Blake • Luca Nanetti • Lisette van der Meer •

Leonardo Cerliani • Remco Renken • Gerdina H. M. Pijnenborg •

Andre Aleman

Received: 16 July 2013 / Accepted: 15 October 2013

� Springer-Verlag Berlin Heidelberg 2013

Abstract Auditory verbal hallucinations (AVH) in

schizophrenia have previously been associated with func-

tional deficiencies in language networks, specifically with

functional disconnectivity in fronto-temporal connections

in the left hemisphere and in interhemispheric connections

between frontal regions. Here, we investigate whether

AVH are accompanied by white matter abnormalities in

tracts connecting the frontal, parietal and temporal lobes,

also engaged during language tasks. We combined diffu-

sion tensor imaging with tract-based spatial statistics and

found white matter abnormalities in patients with schizo-

phrenia as compared with healthy controls. The patients

showed reduced fractional anisotropy bilaterally: in the

anterior thalamic radiation (ATR), body of the corpus

callosum (forceps minor), cingulum, temporal part of the

superior longitudinal fasciculus (SLF) and a small area in

the inferior fronto-occipital fasciculus (IFOF); and in the

right hemisphere: in the visual cortex, forceps major, body

of the corpus callosum (posterior parts) and inferior pari-

etal cortex. Compared to patients without current halluci-

nations, patients with hallucinations revealed decreased

fractional anisotropy in the left IFOF, uncinate fasciculus,

arcuate fasciculus with SLF, corpus callosum (posterior

parts–forceps major), cingulate, corticospinal tract and

ATR. The severity of hallucinations correlated negatively

with white matter integrity in tracts connecting the left

frontal lobe with temporal regions (uncinate fasciculus,

IFOF, cingulum, arcuate fasciculus anterior and long part

and superior long fasciculus frontal part) and in inter-

hemispheric connections (anterior corona radiata). These

findings support the hypothesis that hallucinations in

schizophrenia are accompanied by a complex pattern of

Electronic supplementary material The online version of thisarticle (doi:10.1007/s00429-013-0663-y) contains supplementarymaterial, which is available to authorized users.

B. Curcic-Blake (&) � L. Nanetti � L. van der Meer �L. Cerliani � R. Renken � G. H. M. Pijnenborg � A. Aleman

Department of Neuroscience, Neuroimaging Center (NIC),

University Medical Center Groningen, University of Groningen,

A. Deusinglaan 2, 9713AW Groningen, The Netherlands

e-mail: [email protected]

L. van der Meer

Department of Rehabilitation, Lentis, Zuidlaren,

The Netherlands

L. van der Meer

Rob Giel Research Centrum, University of Groningen,

University Medical Center Groningen, Groningen,

The Netherlands

L. Cerliani

Netherlands Institute for Neuroscience, Royal Netherlands

Academy of Arts and Sciences (KNAW), Amsterdam,

The Netherlands

G. H. M. Pijnenborg

Department of Clinical Psychology & Experimental

Psychopathalogy, University of Groningen, Grote Kruisstraat

2/1, 9712TS Groningen, The Netherlands

G. H. M. Pijnenborg

Department of Psychotic Disorders, GGZ-Drenthe,

Dennenweg 9, 9404LA Assen, The Netherlands

123

Brain Struct Funct

DOI 10.1007/s00429-013-0663-y

white matter alterations that negatively affect the language,

emotion and attention/perception networks.

Keywords Hallucinations � Anatomical

connectivity � Diffusion tensor imaging (DTI) �Language network � Thalamo-cortical connectivity �Fronto-temporal connectivity

Abbreviations

ACC Anterior cingulate cortex

ACR Anterior corona radiata

AF Anterior fasciculus

ATR Anterior thalamic radiation

AVH Auditory verbal hallucinations

DTI Diffusion tensor imaging

IFG Inferior frontal gyrus

IFOF Inferior fronto-occipital fasciculus

ILF Inferior longitudinal fasciculus

FA Fractional anisotropy

ROI Region of interest

SLF Superior longitudinal fasciculus

TBSS Tract-based spatial statistics

TFCE Threshold-free cluster enhancement

UF Uncinate fasciculus

Introduction

Auditory verbal hallucinations (AVH, or ‘‘hearing voices’’)

are distressing symptoms of schizophrenia that affect the

majority of patients and diminish their quality of life (Al-

eman and Larøi 2008). Recent developments in brain

imaging reveal that AVH in schizophrenia are associated

with a complicated brain connectivity dysfunction (Jardri

et al. 2011; Brown and Thompson 2010) that incorporates

both functional and anatomical connections (Allen et al.

2008). Diffusion tensor imaging (DTI) is a relatively novel

non-invasive technique to investigate the integrity of ana-

tomical connections in the brain. This method uses a spe-

cific sequence of MR imaging to estimate white matter

structural integrity as reflected by regional water diffusivity

in white matter tracts (Beaulieu 2002). Although several

studies have used DTI to investigate anatomical connec-

tivity in schizophrenia, only a handful of them have

extended this research to find the underlying associations

with hallucinations. In addition, previous studies have

chiefly focused on the integrity of only one tract the arcuate

fasciculus (AF), neglecting the full network. Even more,

the variety of techniques used in these studies has led to

inconsistent results (Shergill et al. 2007; Catani et al. 2011;

Hubl et al. 2004; de Weijer et al. 2011).

Dysfunction of the language processing network, consti-

tuted of Broca’s and Wernicke’s areas, their right hemisphere

homologs and the anterior cingulate gyrus (ACC), has been

associated with AVH (Allen et al. 2008; Mechelli et al. 2007;

Jardri et al. 2011; Kuhn and Gallinat 2010). Several theories of

hallucinations rely on the idea of a lack of synchronization

between language areas such as Broca’s and Wernicke’s areas

(Grossberg 2000; Aleman and Larøi 2008; Hoffman 2007).

Indeed, in two previous studies we found decreased functional

connectivity among these regions during rest (Vercammen

et al. 2010) and during a phonological language task (Curcic-

Blake et al. 2012). More specifically, when inner speech is

initiated, the strength of the effective connectivity between

Wernicke’s and Broca’s areas is negatively associated with

hallucinations, and the interhemispheric connections from

Broca’s homolog (the equivalent area in the right hemisphere)

to Broca’s area are weakened (Curcic-Blake et al. 2012). In the

present article, we aim at investigating the anatomical

underpinnings of these findings, focusing on pathways con-

necting the regions involved in language processing, namely

fronto-temporo-parietal, bilateral frontal (Broca’s area and the

right hemisphere homolog) and ACC.

Finding the anatomical pathways that connect specific

functional regions is not straightforward. Most available

methods investigate separately either functional activation

(fMRI) or anatomical connections (DTI), but do not com-

bine the two. However, using intraoperative direct electr-

ostimulation, Duffau (2008) and Mandonnet et al. (2007)

were able to distinguish pathways relevant to language

tasks by selectively disrupting particular connections dur-

ing the execution of a task. They identified the inferior

fronto-occipital fasciculus (IFOF), superior longitudinal

fasciculus (SLF), AF and uncinate fasciculus (UF; relayed

to the inferior longitudinal fasciculus) as pathways that are

both chiefly involved in language and in connecting the

frontal and temporo-parietal language areas. Also impor-

tant are the genu of the corpus callosum and the anterior

corona radiate (ACR) which connect the left and right

frontal cortex (Ranson 1947), and the cingulum which

connects the ACC with the posterior brain regions.

We used DTI combined with tract-based spatial statistics

(TBSS) (Smith et al. 2006) to investigate these white matter

pathways in the context of AVH. DTI-derived measures such

as fractional anisotropy (FA) have been proven to reflect the

integrity of the axonal membrane and myelin sheath

(Beaulieu 2002), as well as the local coherence of myelinated

axons (Cercignani and Horsfield 2001). Focusing on FA in

the framework of TBSS for assessing the relationship

between white matter integrity and the presence of halluci-

nations in patients with schizophrenia puts the present study

in line with previous work on neuropsychiatric conditions

(Benedetti et al. 2011).

Brain Struct Funct

123

Methods

Participants

Patients with a DSM-IV diagnosis of schizophrenia and

healthy controls participating in a study of insight into

psychosis (van der Meer et al. 2012) were included in our

DTI study. We selected only right-handed subjects,

because handedness has previously been proven to influ-

ence the brain lateralization and the thickness of fibers

(Parker et al. 2005), but not consistently for left-handed

people (Knecht et al. 2000). To confirm the diagnosis of

patients and the absence of psychiatric disorders in healthy

controls, all subjects were assessed by the Mini Interna-

tional Neuropsychiatric Interview-Plus (Sheehan et al.

1998). The patients were divided into two groups according

to item M6 of this interview: patients who either never

experienced and/or did not have any AVH in the last month

(NoAVH group), and patients who had AVH in the past

month (AVH group). The severity of symptoms was

determined by the Positive and Negative Syndrome Scale

(PANSS) interview (Kay et al. 1987). Two patients were

excluded because their reported presence of AVH was

inconsistent with the severity [PANSS P3 item score of\3,

indicating the absence of hallucinations (Kay et al. 1987),

together with a MINI interview item M6 indicating the

presence of hallucinations]. To clarify this issue, we aimed

to ensure that the patients in the AVH group had halluci-

nations in the auditory verbal modality, regardless of other

modalities (visual or tactile). Similarly, we aimed to ensure

that the NoAVH group had no AVH in the month prior to

the study and some of them had never experienced AVH

(n = 6). Insight into psychosis was assessed using the

Schedule for the Assessment of Insight-Expanded Version

(SAI-E) (Kemp and David 1997). All but two patients were

taking at least one type of antipsychotic medication.

Medication effects were included in additional analysis

according to a standardized quantitative method for com-

paring dosages of different drugs (Andreasen et al. 2010).

Each medication dose was expressed in equivalent doses of

haloperidol.

This study was approved by the Medical Ethical Com-

mittee of the University Medical Hospital in Groningen.

All the participants signed a written informed consent.

They received a monetary compensation (45 Euros) upon

their participation.

Imaging and data analysis

Diffusion-weighted images were acquired using a 3T Phi-

lips Intera MRI scanner (Philips, Best, The Netherlands)

with an 8-channel SENSE head coil using a single-shot

pulsed gradient spin echo EPI sequence (TR = 7,621 ms,

TE = 113 ms, SENSE factor 3, p-reduction, and POS

factor of 1). Fifty-one slices were acquired with the fol-

lowing parameters: field of view 240 9 240 mm; matrix

size 128 9 128; image resolution 1.875 9 1.875 9 2 mm3

without gap. In total, 67 volumes were acquired per sub-

ject, 7 without diffusion weighting (b0 = 0 s/mm2; aver-

aged automatically by the scanner) and 60 volumes with

diffusion weighting (b = 4,000 s/mm2) along 60 isotropi-

cally distributed directions.

Eddy current and motion artifacts were corrected for,

and all diffusion images were aligned to the reference

volume (first b0 image) using the FSL toolbox (http://

www.fmrib.ox.ac.uk/fsl). Subsequently, voxels related to

the brain were separated from the skull using the FSL Brain

Extraction Tool and the diffusion tensor was fitted to the

images using FSL FDT Diffusion to estimate diffusion

tensor parameters and related indices of white matter

integrity (FA). Maps of the FA parameter were fed to the

TBSS processing pipeline (Smith et al. 2006).

The TBSS method involves the normalization of diffu-

sion-weighted images in such a way that the center of the

main tracts is aligned for each subject, where between-

group differences are most informative and less likely due

to partial volume artifacts. The use of this method

decreases the likelihood of observing differences between

groups that arise from partial volume artifacts. This

increases the statistical power of the analysis, enabling the

detection of medium and small effects, which are more

frequent in white matter studies of neuropsychiatric syn-

dromes in comparison to neurological conditions. At this

point, one subject was excluded from further analysis since

its overall FA (within the white matter skeleton) was more

than three standard deviations lower than the average of the

overall FA value within the patient group. FA is very

sensitive to unspecific noise (Bastin et al. 1998; Alexander

et al. 2007); thus, such extreme value is usually indicative

of some unknown measurement artifacts such as move-

ment. We carried out (1) voxelwise and (2) regions of

interest (ROI) analysis, as described below.

Voxelwise comparison

Within the white matter tract skeleton, we estimated the

voxelwise FA differences between controls and schizo-

phrenia patients, as well as within the patient group.

Inference was carried out using the FSL randomize tool for

nonparametric permutation testing: between-group FA

values were compared with the corresponding null distri-

butions estimated by exchanging subject labels 5,000

times. As a standard procedure, the age of the subjects was

entered as an additional regressor. The additional covari-

ates that were determined significantly by ANCOVA (see

below) were entered in comparisons between two patient

Brain Struct Funct

123

groups. The final p threshold of 0.05 was corrected for

multiple comparisons using threshold-free cluster

enhancement (TFCE) (Smith and Nichols 2009), corre-

sponding to an FWE correction (p = 0.05).

ROI analysis and covariates

Masks for ROI (see Fig. 1) were constructed as the inter-

section of the mean FA skeleton with the masks for par-

ticular white matter tracts. Masks for most of the ROIs

were obtained from the following JHU DTI-based white

matter atlases (Wakana et al. 2007): ICBM-DTI-81 white

matter labels for ACR left and right, genu and cingulum;

JHU-white matter tractography atlas for IFOF, UF and

SLF. The AF is not listed in the atlases available with the

FSL software. Therefore, masks for the AF were obtained

from the Natbrainlab atlas for white matter tracts (Thiebaut

de Schotten et al. 2011) consisting of three AF sections: the

anterior, long and posterior AF (Fig. 1). The average FA

per ROI was then calculated for each subject from their FA

skeletons. To determine which covariates to take into

account in the follow-up correlation analysis, ANCOVA

was performed on the averaged FA per ROI with halluci-

nation group (AVH or NoAVH) as a fixed factor, and age,

duration of illness, medication effects and SAI-E scores as

additional explanatory variables. Age is standardly used as

a covariate in white matter tract analysis, because it

strongly influences integrity of white matter (Jones et al.

2006; de Weijer et al. 2011). Duration of illness was

recently acknowledged as an important confounder as there

is evidence that the duration of schizophrenia illness affects

the FA values (Mori et al. 2007; Rotarska-Jagiela et al.

2009). Furthermore, medication was previously reported as

a possible source of influence on the white matter integrity,

Fig. 1 Masks for ROI analysis:

uncinate fasciculus (yellow in

a); anterior corona radiata

(ACR; Left red, right blue);

genu (green); cingulum (light

blue); SLF (green in c); anterior

AF (AFant; dark orange in c,

d); longitudinal AF (AFlong;

light orange in c, d); posterior

AF (AFpost; light pink in c, d);

overlap between SLF and

longitudinal AF (yellow in c);

inferior fronto-occipital

fasciculus (IFOF; pink). Cross

sections, MNI coordinates in

a at z = -8, axial view; b at

z = 0, axial view; c at z = 34,

axial view; d at x = -45

coronal view

Brain Struct Funct

123

although this is still under debate (Kubicki et al. 2005). To

include it as a covariate, for each subject we calculated the

medication quantity equivalent dose of haloperidol

according to Andreasens’s formula for antipsychotic

medication (Andreasen et al. 2010). For nine patients, for

one type of medication (often patients used combined

medication) the dosage was unknown. In those cases, we

averaged quantity for this particular medication and then

calculated the haloperidol equivalent. In four cases,

patients also received antipsychotics not listed in the article

(Andreasen et al. 2010), Effexor (venlafaxine; 39) and

Orap (19). For them, we used averaged medication

quantities across all subjects. Finally, because the patients

were selected according to their insight rating—SAI-E

score, we also entered SAI-E scores in ANCOVA.

The average FA values per subject for each region were

subsequently entered into a partial correlation analysis to

investigate the correlation between the p3 score and ROI

value using SPSS package software. An FDR correction for

multiple comparisons was performed on each p value from

the correlation analysis.

Results

Subjects

The three groups of subjects (healthy controls, schizophrenia

patients with and without current hallucinations) did not sig-

nificantly differ in age, gender, education level, negative

PANSS score or general symptom severity (details in

Table 1). However, the two patient groups differed in their

positive PANSS scores. This was expected as this scale

contains the p3 item for hallucinations. In the group of patients

without current hallucinations, six patients had never experi-

enced AVH before.

White matter integrity differences

between schizophrenia patients and healthy subjects

The results of voxelwise TBSS comparison (p \ 0.05

TFCE corrected) were in accordance with previous reports

(Ellison-Wright and Bullmore 2009). The schizophrenia

patients exhibited bilaterally decreased FA values in the

following regions: anterior thalamic radiation (ATR, site of

the IFOF/UF/ACR); body of the corpus callosum (forceps

minor); cingulum (posterior parts); temporal part of the

SLF and a small area in the IFOF. In the right hemisphere,

schizophrenia patients had decreased FA in the visual

cortex-optic radiation, forceps major, body of the corpus

callosum (posterior parts) and inferior parietal lobule/

auditory cortex. The results are summarized in Fig. 2.

Hallucinations and FA

TBSS tracts of 32 right-handed schizophrenia patients were

entered for the voxelwise and ROI patient comparison. We

first compared patients using the voxelwise TBSS compari-

son (p \ 0.05 TFCE corrected). The confounding effects of

age and duration of illness were accounted for by introducing

their respective scores as covariates. Patients with current

hallucinations had significantly decreased FA on higher

threshold (p \ 0.02 TFCE corrected) in the left hemisphere

(see Fig. 3): anterior IFOF, UF, ACR, AF especially the

anterior and long parts of the AF (branches reaching frontal

regions such as BA44, and temporal regions), callosal body

Table 1 Demographic data of subjects

Mean (SD) Significance

Healthy Schizophrenia

patients no AVH

Schizophrenia

patients AVH

Three groups No AVH vs AVH

(n = 14) (n = 14) (n = 17)

Age in years 34 (11) 34 (12) 35 (11) F(2,42) = 0.02 (0.98)

Gender males 9 11 13 v2(2,42) = 0.87 (0.65)

Education 5.7 (1.0) 5.2 (1.1) 5.2 (1.2) F(2,42) = 0.7 (0.35)

PANSS pos. 11.4 (4.4) 16.9 (4.0) U = 38 (0.001)

PANSS neg. 13.9 (4.5) 12.8 (4.3) U = 102 (0.19)

PANSS gen. 27.7 (4.7) 29.7 (6.4) U = 99 (0.43)

PANSS tot. 53.1 (9.4) 59.4 (12.0) U = 82 (0.14)

Insight 14.4 (5.4) 11.6 (6.5) T(29) = 1.3 (0.20)

Duration of illness in years 12.6 (8.0) 15.3 (10.4) T(29) = -0.8 (0.44)

Medication (mg) haloperidol equivalent 6.9 (5.7) 7.4 (4.8) U = 102.5 (0.51)

The left column lists the demographic variables. The 2nd–4th columns from the left show average values of the variables across the group, with their

standard deviations in brackets. Education level was rated according to a six-point scale defined by Verhage, which ranges from primary school (1) to

university level (6). Nonparametric tests were used to test the group difference for PANSS (Mann–Whitney test) and gender (Chi square)

Brain Struct Funct

123

(medial and posterior part—the forceps major), the cingu-

lum, corticospinal tract and the ATR. In addition, on a lower

threshold (p \ 0.05 TFCE corrected) patients with current

hallucinations had decreased FA in several regions in the

right hemisphere including the anterior part of AF, parts of

SLF, ACR, cingulum and the corticospinal tract.

Covariate analysis revealed that age affected the FA in

IFOF and anterior AF and only marginally in other tracts (see

Online Resource 1). However, the duration of illness was

significant for several tracts: the anterior AF, the IFOF, the

cingulum and posterior and long portions of the AF (trend).

Insight and medication revealed no effect on the FA. There-

fore in addition to the standard age covariate, we also included

the duration of illness in all our analysis. Post hoc contrast of

group differences between patients with and without current

hallucinations revealed a significant decrease in FA in the left

IFOF (F(1,27) = 10.85; pFDR = 0.027), left ACR

(F(1,27) = 847; pFDR = 0.036), left UF (F(1,27) = 7.84;

pFDR = 0.021), left cingulum (F(1,27) = 6.78; pFDR =

0.037), anterior part of AF (F(1,27) = 6.35; pFDR = 0.036),

posterior part of AF (F(1,27) = 5.81; pFDR = 0.038) and SLF

(F(1,27) = 5.79; pFDR = 0.033).

Partial correlation ROI analysis (right-handed hallucina-

tions patients, n = 17; no hallucinations, n = 14) revealed a

negative correlation of FA and P3 in the left IFOF, UF,

cingulum, left and right ACR, anterior and long part of AF

and SLF (see Table 2 for r values and significance). We

found that the association between AF impairment and hal-

lucination severity is most evident in the anterior part of the

AF. Details of partial correlation analysis and FDR correc-

tion calculation are given in Online Resource 2. The distri-

bution of FA values in the ROIs is depicted in Fig. 4.

Discussion

We found that alterations of white matter in tracts that

connect frontal and temporal areas (especially in the left

Fig. 2 Results of voxelwise

analysis: comparison of healthy

controls and schizophrenia

patients. In figure panels the

following are depicted: ATR

left/anterior limb of internal

capsule (z = -3 and z = 5);

IFOF (z = 5 and x = -24);

forceps minor/genu (z = 10 and

z = 17); forceps major

(z = 17); cingulum (z = 17 and

x = -16 and x = -12); corpus

callosum (z = 30 and x = -6

and x = 2 and x = 27); SLF

right (z = 30 and x = -12);

primary auditory cortex (z = 5);

optic radiation (x = 27)

Brain Struct Funct

123

hemisphere), as well as in tracts involved in interhemi-

spheric communication, are associated with hallucination

severity in schizophrenia patients. Anatomical impairment

does not merely correspond to the presence/absence of

hallucinations. Rather, the extent of the anatomical

impairment is directly correlated with the severity of the

hallucinations. More specifically, patients with hallucina-

tions had smaller FA values than patients without current

hallucinations in tracts connecting frontal and temporal

areas involved in language (the uncinate fasciculus, IFOF,

the SLF and anterior and long arcuate fasciculus), as well

as in tracts involved in interhemispheric communication,

namely the bilateral ACR and posterior parts of the corpus

callosum. The finding of fronto-temporal and

interhemispherical deficiencies (see Fig. 3; Table 2) sup-

ports theories in which hallucinations arise from the erro-

neous attribution of internally generated information

(Friston and Frith 1995) such as inner speech, accompanied

by disrupted fronto-temporal connections between lan-

guage processing areas (Allen et al. 2008).

FA is a measure of the overall directionality of water

diffusion and is largest when there is a clear diffusion

direction (such as in white matter tracts) and lowest in

media where there is little restriction to diffusion (such as

cerebrospinal fluid). Thus, a decrease in FA values between

groups reflects diminished connections, most probably due

to the orientation or number of axons, or reduced integrity

of the axonal or myelin sheath (Beaulieu 2002). It has

previously been argued that decreased FA reflects a

decrease in anatomical connectivity (Rotarska-Jagiela et al.

2009). We found not only that all schizophrenia patients

exhibited significantly decreased FA values in comparison

to healthy controls, but also that this decrease became more

pronounced as the hallucinations score increased. This

suggests that the anatomical connections are disrupted even

more in patients with hallucinations. The abnormalities in

the fronto-parietal tracts and interhemispheric tract

involved in language processing (Fig. 2) are consistent

with our previous investigation of functional connectivity

in the language network (Curcic-Blake et al. 2012). We

found specifically that during a phonological task, patients

with AVH showed significantly reduced connectivity from

Wernicke’s to Broca’s areas and a trend toward a reduction

in connectivity from the homologs of Broca’s and

Fig. 3 Results of voxelwise

analysis: comparison of

hallucinating and non-

hallucinating patients. The

following are depicted: z = -6:

the anterior part of IFOF;

z = -5: the UF; z = 20: the

posterior part of IFOF, anterior

thalamic radiation and SLF/AF

branching frontal; z = 33: the

cingulum and the SLF/AF

branching temporal; y = -45:

posterior callosal body; y = 15:

the callosal body, SLF, anterior

thalamic radiation, the anterior

part of IFOF and UF; y = -20:

corticospinal tract and SLF

temporal

Table 2 Results of ROI analysis

Tract Correlation p pFDR

IFOF -0.510 0.005 0.047

UF -0.497 0.006 0.030

Cing -0.493 0.007 0.022

ACRl -0.482 0.008 0.020

ACRr -0.453 0.014 0.027

AFant -0.452 0.014 0.023

SLF -0.423 0.022 0.032

AFlong -0.413 0.026 0.033

AFpost -0.245 0.200 0.223

Genu -0.185 0.337 0.337

The mean FA values per subject were correlated with the p3 scores

after the effects of age and illness duration were accounted for by

means of partial correlations

Brain Struct Funct

123

Wernicke’s areas to Broca’s area in the right hemisphere.

In other words, patients without AVH had intermediate

functional connectivity strengths, similar to the anatomical

connectivity strengths found in the current study.

Our finding of diminished connectivity between frontal

and temporo-parietal regions also yields information that

may complement the so-called corollary discharge

hypothesis (Feinberg and Guazzelli 1999; Frith 2005). This

hypothesis states that the improper source attribution of

voices in patients with hallucinations can be explained in

terms of insufficient top-down feedback and control from

higher cognitive areas (such as Broca’s region) toward the

primary and secondary auditory areas (such as Wernicke’s

region). Indeed, recent experimental studies point toward a

delay in this top-down control (Hoffman et al. 2011;

Whitford et al. 2011). Consequently, our findings may

suggest that abnormal fronto-temporoparietal connections

in patients with hallucinations are responsible for the

decreased functional connectivity between these areas, a

putative underlying mechanism of reduced corollary

discharge.

Two fasciculi of the left hemisphere, the IFOF and the

UF, both belong to the language network and were found

significantly altered in schizophrenia patients. The IFOF

bundle connects the frontal and temporal regions, more

specifically the frontal lobe with the posterior portion of the

occipital gyri via the posterior temporo-basal area (Martino

et al. 2010). The IFOF has only recently been identified as

an important connection for the semantic processing of

language (Duffau 2008). Next to the AF, which is involved

in phonological language processing, Duffau describes the

IFOF as the ventral route for language processing. A clear

anatomical, also DTI-oriented, description of this impor-

tant connection appeared only recently in literature (Forkel

et al. 2012), in which a thorough literature review was

combined with DTI tractography of the IFOF. IFOF also

has an important role in primary visual processing (ffytche

and Catani 2005), visual imagery, such as visual halluci-

nations (Chechlacz et al. 2012) or visualization of music

(Zamm et al. 2013), visual awareness (Amad et al. 2013),

and processing of emotions (Baggio et al. 2012; Xu et al.

2012). Specifically, Forkel and colleagues outlined how

IFOF connects two branches of the frontal lobe, namely the

inferior frontal gyrus (where Broca’s area is placed) and

the medial frontal gyrus. They further described the route

of this fasciculus as ventral, passing the temporal lobe and

ending in the occipital lobe. Although previous studies

found indications that the IFOF plays a role in schizo-

phrenia (Clark et al. 2011), the direct involvement of the

IFOF in AVH is investigated for the first time in our cur-

rent work.

The left UF is also a pathway that connects the temporal

regions and the frontal lobe, in particular the inferior

frontal gyrus. The UF is traditionally considered to be part

of the limbic lobe (Von Der Heide et al. 2013) and its

function is not entirely clear. Recently, it has been found

that the UF is involved in the semantic processing of lan-

guage (Catani et al. 2013), an alternative route to the IFOF

for language processing (Duffau et al. 2009). However, the

role of the UF is still under debate, with evidence

Fig. 4 Block diagram of FA

values in ROI for hallucinating

(dark gray) and non-

hallucinating group (light gray).

Significant differences revealed

by post hoc analysis of

ANCOVA (the age and duration

of illness were accounted for)

are indicated by asterisk

Brain Struct Funct

123

suggesting that it acts as a pathway for creating mnemonic

associations involving the processing of language, visual

and auditory information (Von Der Heide et al. 2013). Our

findings of abnormalities in the left UF in schizophrenia

patients compared to healthy controls are consistent with

the results of Hubl et al. (2004) and de Weijer et al. (2011).

Our finding that the degree of FA reduction in the UF is

associated with the severity of hallucinations is, however,

an original and novel finding of the present study.

We found a negative correlation between FA in both the

SLF and the anterior and long part of AF, and the severity

of hallucinations. The SLF is a complex association tract

that links multiple frontal, temporo-parietal and occipital

regions (Makris et al. 2005). It is considered the most

important fasciculus in language processing because it

provides direct connections between the language regions

Broca’s and Wernicke’s areas. According to Makris et al.

(2005), the AF, which is often investigated in studies of

auditory hallucinations, is part of the SLF originating in the

dorsal prefrontal cortex (Brodmann areas 6 and 46), pass-

ing around the caudal Sylvian fissure and ending in the

caudal area of the superior temporal sulcus.

Our results are in agreement with the findings of Catani

et al. (2011) who performed exhaustive tractography ana-

lysis on the AF and observed that schizophrenia patients

with hallucinations showed reduced FA in the long seg-

ment of the left AF consistent with our findings. They were

able to examine the details of fibers within the AF tract and

to pinpoint the posterior temporal and anterior regions in

the inferior frontal and parietal lobes where the differences

occurred. However, the findings of this study are not

entirely consistent with the previous literature. For exam-

ple, Shergill et al. (2007) reported a positive association of

FA along the AF with hallucinations, where Hubl et al.

(2004) found both increased and decreased FA values in

the left AF when patients with and without hallucinations

were compared [evident from Fig. 5a (1 and 2) in their

article]. On the other hand, de Weijer et al. (2011) did not

find any direct correlation between the severity of hallu-

cinations and the FA values in the left AF after excluding

the influence of age. However, they found that the mag-

netic transfer ratio (MTR) was increased in patients with

hallucinations, while FA was negatively correlated with

MTR, thus supporting our findings.

The reasons for these discrepancies might lie in the

different data acquisition methods used in the above studies

(line-by-line spin echo sampling by Hubl et al.), the dif-

ference in analysis methods and the different question-

naires (BPRS was used by Shergill et al.). Another

important aspect is that the duration of illness may partially

explain the reduced FA in schizophrenia patients (Mori

et al. 2007). Indeed, Rotarska-Jagiela et al. (2009) found

positive correlations of the duration of illness with the FA

values in the bilateral arcuate fasciculus and negative

correlations in the corpus callosum, occipitofrontal fas-

ciculus and other areas.

The ACR and the posterior corpus callosum are also the

loci of some significant differences between hallucinating

and non-hallucinating schizophrenia patients. The ACR

consists of a mixture of various association, projection and

callosal fibers, and partially connects the corpus callosum

with the cortex (Wakana et al. 2004), thus providing an

extension to the interhemispheric connection. It has been

implied that the ACR connects the ACC with other brain

structures (Tang et al. 2010) and is involved in attention

networks (Niogi et al. 2010). We found an association

between the bilateral ACR and hallucination severity. The

majority of the left ACR mask overlapped with the dif-

ferences between schizophrenia patients and healthy con-

trols, suggesting that these differences in the left

hemisphere are not only a characteristic of schizophrenia,

but also to some extent of hallucinations. However, the

differences observed in the right ACR are specific for

hallucinations. Similarly, we found significant differences

in the voxelwise comparison of the two patient groups in

the posterior parts of the corpus callosum. These findings

are consistent with that of Knochel et al. (2012) who

reported that hypoconnectivity in the corpus callosum was

associated with the severity of hallucinations.

Similarly, consideration can be made about the cingu-

lum, which is related to the anterior cingulate (also

involved in attention and sensory decision making, as

mentioned above). An impeded connection along the left

cingulum might contribute to the overall impedance of this

sensory processing. The ACC provides ‘the interface for

emotions’; it is known to be important in schizophrenia

(Barch and Dowd 2010). In line with these findings, Ver-

cammen et al. revealed a reduction in coupling between the

temporo-parietal junction and ACC in patients with AVH.

We found decreased FA values in the left cingulate of

patients in the hallucination group compared to patients

without current hallucinations, a result that is in agreement

with Hubl et al. (2004). These studies are consistent with

theories of decreased attribution of agency, and thus

decreased awareness of one’s own volitional actions (in

this case inner speech production) in AVH patients (Frith

1996).

A number of limitations of our study deserve discussion.

First, besides acting as an indicator of white matter integ-

rity, FA might be influenced by other factors such as the

thickness of fibers (Beaulieu 2002). However, in our TBSS

analysis the FA values were extracted from regions

belonging to a mean estimated skeleton mask. Thus, vari-

ations in fiber thickness beyond the skeleton do not affect

our results. This also implies that our method is unsuitable

for the investigation of fiber thickness. Second, DTI does

Brain Struct Funct

123

not differentiate between efferent and afferent projections,

thus it does not provide any information about the causality

and directionality of the connections. Third, we investi-

gated differences between patients with and without cur-

rent hallucinations according to the PANSS and MINI

interviews. Patients were considered to be without current

hallucinations if none were reported in the month prior to

scanning. However, they may have experienced halluci-

nations earlier in their illness, and thus may still have had

trait characteristics associated with hallucinations. This

implies that our conclusions are limited to the severity of

hallucinations rather than their presence/absence, espe-

cially because schizophrenia patients who have never

experienced hallucinations are difficult to find. Fourth, we

did correct for the dose of medication in schizophrenia

patients, rounding the doses that were missing for nine

patients. However, all these patients were medicated and

we used a standard method for filling in unknown param-

eters (by means of specific mean values). This is a rec-

ommended method because it minimizes bias while

maintaining the df as compared to simply omitting the data

(Rubin et al. 2007). Furthermore, a previous study (Seok

et al. 2007) found that controlling for medication dosage

did not change the correlation between FA values in the

SLF and the hallucination severity, in line with our findings

that medication dosage did not affect the FA values in any

ROI.

All the fasciculi investigated in this paper are considered

to be involved in language processing. However, their

function is often much broader and encompasses the pro-

cessing of emotions (cingulate, ACR), visual information

(IFOF) or episodic memory (UF). Consequently, AVHs are

also associated with broader functional deficiencies, such

as cognitive control and emotional processing (Aleman and

Larøi 2008). Future studies should investigate the role of

these regions in more detail. For example, the AF seems to

be more asymmetric in men (Catani and Mesulam 2008)

and the lateralization of the UF is still under debate

(Highley et al. 2002; Thiebaut de Schotten et al. 2011).

Therefore, an important future strategy will be to combine

the volumetric asymmetries of these regions in relation to

language.

To conclude, our results suggest that hallucinations in

schizophrenia patients are associated with a complex set of

white matter abnormalities that involve at least three dis-

tinct systems: the language network, putatively in the

‘inner speech’ domain; and the central in attention and

perception loop; and the limbic system, providing the

emotional edge. Additionally, we advance the hypothesis

that the severity of the hallucinations experienced by

schizophrenia patients is a direct function of the extent of

the impairment affecting the anatomical connectivity

within these three systems.

Acknowledgments The authors thank G.R. Blake for his comments

on earlier versions of the manuscript. We thank psychiatrists Dr.

R. Bruggeman and Dr. H. Knegtering for their help with patient

inclusion. Our thanks are due especially to A. de Vos and J. van der

Velde for help with data sorting and to S. Chalavi for discussions on

analysis. L. Cerliani is supported by an NWO MaGW Open compe-

tition grant (grant number 400-08-089) and by an NIHC grant (grant

number 056-13-017). This research was supported by a EURYI

Award from the European Science Foundation (No. 044035001)

awarded to A.A.

Conflict of interest The authors declare that they have no com-

peting financial interests in relation to the work described.

References

Aleman A, Larøi F (2008) Hallucinations: the science of idiosyncratic

perception. American Psychological Association, Washington,

DC

Alexander AL, Lee JE, Lazar M, Field AS (2007) Diffusion tensor

imaging of the brain. Neurotherapeutics 4:316–329

Allen P, Larøi F, McGuire PK, Aleman A (2008) The hallucinating

brain: a review of structural and functional neuroimaging studies

of hallucinations. Neurosci Biobehav Rev 32:175–191

Amad A, Cachia A, Gorwood P, Pins D, Delmaire C, Rolland B et al

(2013) The multimodal connectivity of the hippocampal com-

plex in auditory and visual hallucinations. Mol Psychiatry (in

press)

Andreasen NC, Pressler M, Nopoulos P, Miller D, Ho BC (2010)

Antipsychotic dose equivalents and dose-years: a standardized

method for comparing exposure to different drugs. Biol Psychi-

atry 67:255–262

Baggio HC, Segura B, Ibarretxe-Bilbao N, Valldeoriola F, Marti MJ,

Compta Y et al (2012) Structural correlates of facial emotion

recognition deficits in Parkinson’s disease patients. Neuropsych-

ologia 50:2121–2128

Barch DM, Dowd EC (2010) Goal representations and motivational

drive in schizophrenia: the role of prefrontal–striatal interactions.

Schizophr Bull 36:919–934

Bastin ME, Armitage PA, Marshall I (1998) A theoretical study of the

effect of experimental noise on the measurement of anisotropy in

diffusion imaging. Magn Reson Imaging 16:773–785

Beaulieu C (2002) The basis of anisotropic water diffusion in the

nervous system—a technical review. NMR Biomed 15:435–455

Benedetti F, Yeh PH, Bellani M, Radaelli D, Nicoletti MA, Poletti S,

Falini A, Dallaspezia S, Colombo C, Scotti G, Smeraldi E,

Soares JC, Brambilla P (2011) Disruption of white matter

integrity in bipolar depression as a possible structural marker of

illness. Biol Psychiatry 69:309–317

Brown GG, Thompson WK (2010) Functional brain imaging in

schizophrenia: selected results and methods. Curr Top Behav

Neurosci 4:181–214

Catani M, Craig MC, Forkel SJ, Kanaan R, Picchioni M, Toulopoulou

T, Shergill S, Williams S, Murphy DG, McGuire P (2011)

Altered integrity of perisylvian language pathways in schizo-

phrenia: relationship to auditory hallucinations. Biol Psychiatry

70:1143–1150

Catani M, Mesulam M (2008) The arcuate fasciculus and the

disconnection theme in language and aphasia: history and

current state. Cortex 44:953–961

Catani M, Mesulam MM, Jakobsen E, Malik F, Martersteck A,

Wieneke C et al (2013) A novel frontal pathway underlies verbal

fluency in primary progressive aphasia. Brain 136:2619–2628

Brain Struct Funct

123

Cercignani M, Horsfield MA (2001) The physical basis of diffusion-

weighted MRI. J Neurol Sci 186(Suppl 1):S11–S14

Chechlacz M, Rotshtein P, Hansen PC, Riddoch JM, Deb S,

Humphreys GW (2012) The neural underpinnings of simultan-

agnosia: disconnecting the visuospatial attention network.

J Cogn Neurosci 24:718–735

Clark KA, Nuechterlein KH, Asarnow RF, Hamilton LS, Phillips OR,

Hageman NS, Woods RP, Alger JR, Toga AW, Narr KL (2011)

Mean diffusivity and fractional anisotropy as indicators of disease

and genetic liability to schizophrenia. J Psychiatr Res 45:980–988

Curcic-Blake B, Liemburg E, Vercammen A, Swart M, Knegtering H,

Bruggeman R, Aleman A (2012) When Broca goes uninformed:

reduced information flow to Broca’s area in schizophrenia

patients with auditory hallucinations. Schizophr Bull 39(5):

1087–1095

de Weijer AD, Mandl RC, Diederen KM, Neggers SF, Kahn RS,

Hulshoff Pol HE, Sommer IE (2011) Microstructural alterations

of the arcuate fasciculus in schizophrenia patients with frequent

auditory verbal hallucinations. Schizophr Res 130:68–77

Duffau H (2008) The anatomo-functional connectivity of language

revisited. New insights provided by electrostimulation and

tractography. Neuropsychologia 46:927–934

Duffau H, Gatignol P, Moritz-Gasser S, Mandonnet E (2009) Is the

left uncinate fasciculus essential for language? A cerebral

stimulation study. J Neurol 256:382–389

Ellison-Wright I, Bullmore E (2009) Meta-analysis of diffusion tensor

imaging studies in schizophrenia. Schizophr Res 108:3–10

Feinberg I, Guazzelli M (1999) Schizophrenia—a disorder of the

corollary discharge systems that integrate the motor systems of

thought with the sensory systems of consciousness. Br J

Psychiatry 174:196–204

ffytche DH, Catani M (2005) Beyond localization: from hodology to

function. Philos Trans R Soc Lond B Biol Sci 360:767–779

Forkel SJ, Thiebaut de SM, Kawadler JM, Dell’Acqua F, Danek A,

Catani M (2012) The anatomy of fronto-occipital connections

from early blunt dissections to contemporary tractography.

Cortex. doi:10.1016/j.cortex.2012.09.005

Friston KJ, Frith CD (1995) Schizophrenia: a disconnection syn-

drome? Clin Neurosci 3:89–97

Frith C (1996) Neuropsychology of schizophrenia: what are the

implications of intellectual and experimental abnormalities for

the neurobiology of schizophrenia? Br Med Bull 52:618–626

Frith C (2005) The neural basis of hallucinations and delusions. C R

Biol 328:169–175

Grossberg S (2000) How hallucinations may arise from brain

mechanisms of learning, attention, and volition. J Int Neuropsy-

chol Soc 6:583–592

Highley JR, Walker MA, Esiri MM, Crow TJ, Harrison PJ (2002)

Asymmetry of the uncinate fasciculus: a post-mortem study of

normal subjects and patients with schizophrenia. Cereb Cortex

12:1218–1224

Hoffman RE (2007) A social deafferentation hypothesis for induction

of active schizophrenia. Schizophr Bull 33:1066–1070

Hoffman RE, Pittman B, Constable RT, Bhagwagar Z, Hampson M

(2011) Time course of regional brain activity accompanying

auditory verbal hallucinations in schizophrenia. Br J Psychiatry

198:277–283

Hubl D, Koenig T, Strik W, Federspiel A, Kreis R, Boesch C, Maier

SE, Schroth G, Lovblad K, Dierks T (2004) Pathways that make

voices: white matter changes in auditory hallucinations. Arch

Gen Psychiatry 61:658–668

Jardri R, Pouchet A, Pins D, Thomas P (2011) Cortical activations

during auditory verbal hallucinations in schizophrenia: a coordi-

nate-based meta-analysis. Am J Psychiatry 168:73–81

Jones DK, Catani M, Pierpaoli C, Reeves SJ, Shergill SS, O’Sullivan M,

Golesworthy P, McGuire P, Horsfield MA, Simmons A, Williams

SC, Howard RJ (2006) Age effects on diffusion tensor magnetic

resonance imaging tractography measures of frontal cortex

connections in schizophrenia. Hum Brain Mapp 27:230–238

Kay SR, Fiszbein A, Opler LA (1987) The positive and negative

syndrome scale (PANSS) for schizophrenia. Schizophr Bull

13:261–276

Kemp R, David AS (1997) Insight and compliance. In: Blackwell B

(ed) Compliance and the treatment alliance in serious mental

illness. Hardwood Academic Publishers, Amsterdam, pp 61–86

Knecht S, Drager B, Deppe M, Bobe L, Lohmann H, Floel A,

Ringelstein EB, Henningsen H (2000) Handedness and hemi-

spheric language dominance in healthy humans. Brain 123(Pt

12):2512–2518

Knochel C, Oertel-Knochel V, Schonmeyer R, Rotarska-Jagiela A,

van de Ven V, Prvulovic D, Haenschel C, Uhlhaas P, Pantel J,

Hampel H, Linden DE (2012) Interhemispheric hypoconnectiv-

ity in schizophrenia: fiber integrity and volume differences of the

corpus callosum in patients and unaffected relatives. NeuroIm-

age 59:926–934

Kubicki M, Westin CF, McCarley RW, Shenton ME (2005) The

application of DTI to investigate white matter abnormalities in

schizophrenia. Ann N Y Acad Sci 1064:134–148

Kuhn S, Gallinat J (2010) Quantitative meta-analysis on state and

trait aspects of auditory verbal hallucinations in schizophrenia.

Schizophr Bull 32:358–365

Makris N, Kennedy DN, McInerney S, Sorensen AG, Wang R,

Caviness VS Jr, Pandya DN (2005) Segmentation of subcom-

ponents within the superior longitudinal fascicle in humans: a

quantitative, in vivo, DT-MRI study. Cereb Cortex 15:854–869

Mandonnet E, Nouet A, Gatignol P, Capelle L, Duffau H (2007) Does

the left inferior longitudinal fasciculus play a role in language? A

brain stimulation study. Brain 130:623–629

Martino J, Brogna C, Robles SG, Vergani F, Duffau H (2010) Anatomic

dissection of the inferior fronto-occipital fasciculus revisited in the

lights of brain stimulation data. Cortex 46:691–699

Mechelli A, Allen P, Amaro E, Fu CHY, Williams SCR, Brammer

MJ, Johns LC, McGuire PK (2007) Misattribution of speech and

impaired connectivity in patients with auditory verbal halluci-

nations. Hum Brain Mapp 28:1213–1222

Mori T, Ohnishi T, Hashimoto R, Nemoto K, Moriguchi Y, Noguchi

H, Nakabayashi T, Hori H, Harada S, Saitoh O, Matsuda H,

Kunugi H (2007) Progressive changes of white matter integrity

in schizophrenia revealed by diffusion tensor imaging. Psychi-

atry Res 154:133–145

Niogi S, Mukherjee P, Ghajar J, McCandliss BD (2010) Individual

differences in distinct components of attention are linked to

anatomical variations in distinct white matter tracts. Front

Neuroanat 4:2

Parker GJ, Luzzi S, Alexander DC, Wheeler-Kingshott CA, Ciccarelli

O, Lambon Ralph MA (2005) Lateralization of ventral and

dorsal auditory-language pathways in the human brain. Neuro-

image 24:656–666

Ranson SW (1947) The anatomy of the nervous system: its develop-

ment and function. W. B. Saunders company, London

Rotarska-Jagiela A, Oertel-Knoechel V, DeMartino F, van de Ven V,

Formisano E, Roebroeck A, Rami A, Schoenmeyer R, Haenschel

C, Hendler T, Maurer K, Vogeley K, Linden DE (2009)

Anatomical brain connectivity and positive symptoms of

schizophrenia: a diffusion tensor imaging study. Psychiatry

Res 174:9–16

Rubin LH, Witkiewitz K, Andre JS, Reilly S (2007) Methods for

handling missing data in the behavioral neurosciences: don’t

throw the baby rat out with the bath water. J Undergrad Neurosci

Educ 5:A71–A77

Seok JH, Park HJ, Chun JW, Lee SK, Cho HS, Kwon JS, Kim JJ

(2007) White matter abnormalities associated with auditory

Brain Struct Funct

123

hallucinations in schizophrenia: a combined study of voxel-

based analyses of diffusion tensor imaging and structural

magnetic resonance imaging. Psychiatry Res 156:93–104

Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs J, Weiller

E, Hergueta T, Baker R, Dunbar GC (1998) The Mini-

International Neuropsychiatric Interview (M.I.N.I.): the devel-

opment and validation of a structured diagnostic psychiatric

interview for DSM-IV and ICD-10. J Clin Psychiatry 59(Suppl

20):22–33

Shergill SS, Kanaan RA, Chitnis XA, O’Daly O, Jones DK, Frangou

S, Williams SC, Howard RJ, Barker GJ, Murray RM, McGuire P

(2007) A diffusion tensor imaging study of fasciculi in

schizophrenia. Am J Psychiatry 164:467–473

Smith SM, Nichols TE (2009) Threshold-free cluster enhancement:

addressing problems of smoothing, threshold dependence and

localisation in cluster inference. NeuroImage 44:83–98

Smith SM, Jenkinson M, Johansen-Berg H, Rueckert D, Nichols TE,

Mackay CE, Watkins KE, Ciccarelli O, Cader MZ, Matthews

PM, Behrens TE (2006) Tract-based spatial statistics: voxelwise

analysis of multi-subject diffusion data. NeuroImage

31:1487–1505

Tang YY, Lu Q, Geng X, Stein EA, Yang Y, Posner MI (2010) Short-

term meditation induces white matter changes in the anterior

cingulate. Proc Natl Acad Sci USA 107:15649–15652

Thiebaut de Schotten M, Ffytche DH, Bizzi A, Dell’Acqua F, Allin

M, Walshe M, Murray R, Williams SC, Murphy DG, Catani M

(2011) Atlasing location, asymmetry and inter-subject variability

of white matter tracts in the human brain with MR diffusion

tractography. Neuroimage 54:49–59

van der Meer L, de Vos AE, Stiekema AP, Pijnenborg GH, van Tol

MJ, Nolen WA, David AS, Aleman A (2012) Insight in

schizophrenia: involvement of self-reflection networks? Schiz-

ophr Bull 39(6):1288–1295. doi:10.1093/schbul/sbs122

Vercammen A, Knegtering H, den Boer JA, Liemburg EJ, Aleman A

(2010) Auditory hallucinations in schizophrenia are associated

with reduced functional connectivity of the temporo-parietal

area. Biol Psychiatry 67:912–918

Von Der Heide RJ, Skipper LM, Klobusicky E, Olson IR (2013)

Dissecting the uncinate fasciculus: disorders, controversies and a

hypothesis. Brain 136:1692–1707

Wakana S, Jiang H, Nagae-Poetscher LM, van Zijl PC, Mori S (2004)

Fiber tract-based atlas of human white matter anatomy. Radi-

ology 230:77–87

Wakana S, Caprihan A, Panzenboeck MM, Fallon JH, Perry M, Gollub

RL, Hua K, Zhang J, Jiang H, Dubey P, Blitz A, van Zijl P, Mori S

(2007) Reproducibility of quantitative tractography methods

applied to cerebral white matter. Neuroimage 36:630–644

Whitford TJ, Mathalon DH, Shenton ME, Roach BJ, Bammer R,

Adcock RA, Bouix S, Kubicki M, De Siebenthal J, Rausch AC,

Schneiderman JS, Ford JM (2011) Electrophysiological and

diffusion tensor imaging evidence of delayed corollary dis-

charges in patients with schizophrenia. Psychol Med 41:959–969

Xu J, Kober H, Carroll KM, Rounsaville BJ, Pearlson GD, Potenza

MN (2012) White matter integrity and behavioral activation in

healthy subjects. Hum Brain Mapp 33:994–1002

Zamm A, Schlaug G, Eagleman DM, Loui P (2013) Pathways to

seeing music: enhanced structural connectivity in colored-music

synesthesia. Neuroimage 74:359–366

Brain Struct Funct

123