Regionally specific white matter disruptions of fornix and cingulum in schizophrenia.
ABSTRACT Limbic circuitry disruptions have been implicated in the psychopathology and cognitive deficits of schizophrenia, which may involve white matter disruptions of the major tracts of the limbic system, including the fornix and the cingulum. Our study aimed to investigate regionally specific abnormalities of the fornix and cingulum in schizophrenia using diffusion tensor imaging (DTI). We determined the fractional anisotropy (FA), radial diffusivity (RD), and axial diffusivity (AD) profiles along the fornix and cingulum tracts using a fibertracking technique and a brain mapping algorithm, the large deformation diffeomorphic metric mapping (LDDMM), in the DTI scans of 33 patients with schizophrenia and 31 age-, gender-, and handedness-matched healthy controls. We found that patients with schizophrenia showed reduction in FA and increase in RD in bilateral fornix, and increase in RD in left anterior cingulum when compared to healthy controls. In addition, tract-based analysis revealed specific loci of these white matter differences in schizophrenia, that is, FA reductions and AD and RD increases occur in the region of the left fornix further from the hippocampus, FA reductions and RD increases occur in the rostral portion of the left anterior cingulum, and RD and AD increases occur in the anterior segment of the left middle cingulum. In patients with schizophrenia, decreased FA in the specific loci of the left fornix and increased AD in the right cingulum adjoining the hippocampus correlated with greater severity of psychotic symptoms. These findings support precise disruptions of limbic-cortical integrity in schizophrenia and disruption of these structural networks may contribute towards the neural basis underlying the syndrome of schizophrenia and clinical symptomatology.
- SourceAvailable from: Martha E Shenton[show abstract] [hide abstract]
ABSTRACT: Thirty patients with chronic schizophrenia and 30 age-matched controls performed the Attention Network Test (ANT). A subset of the patient group (n=18) also had available magnetic resonance diffusion tensor imaging (DTI) measures of the cingulum bundle (CB) fractional anisotropy and volume. The patients showed a significantly different pattern of ANT performance, characterized primarily by decreased alerting efficiency. In addition, left CB fractional anisotropy correlated significantly with orienting of attention. Smaller right CB volume also correlated with reduced alertness, but not when covarying for medication and illness duration.Schizophrenia Research 03/2007; 90(1-3):308-15. · 4.59 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Evidence suggests that a disruption in limbic system network integrity and, in particular, the cingulate gyrus may play a role in the pathophysiology of schizophrenia. The cingulum bundles (CBs; posterior and anterior) are the most prominent white matter tracts in the limbic system, furnishing both input and output to the cingulate gyrus . In previous diffusion tensor imaging (DTI) studies, abnormal integrity has been demonstrated in the anterior CB portion, but not the posterior, in schizophrenia. As well, the relationships between the abnormalities of CB integrity and the psychopathology of schizophrenia remain to be elucidated. Using DTI acquired on a 3 T MRI machine, we examined fractional anisotropy (FA) in the anterior and posterior CBs of 42 patients with schizophrenia and 24 group-matched controls. Moreover, we investigated the relationships between CB abnormalities and the psychopathology of schizophrenia. Bilaterally reduced FA was demonstrated in both anterior and posterior CBs in schizophrenia patients. However, the pattern of FA reduction was different between anterior and posterior CBs: the reduction in FA was left-accentuated in anterior CBs, while no such lateralized abnormality was found in posterior ones. Finally, FA in posterior CBs correlated with positive symptom scores in patients with schizophrenia. These findings suggest that CB abnormalities in schizophrenia are not restricted to the anterior CB, but include the posterior as well. Pathology in the posterior CB would be one of the possible neural underpinnings of positive symptoms in schizophrenia.Schizophrenia Research 10/2007; 95(1-3):215-22. · 4.59 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: The course and composition of the cingulum bundle was examined by using the autoradiographic tracer technique in the rhesus monkey. The cingulum bundle is observed to consist of three major fiber components originating from thalamus, cingulate gyrus, and cortical association areas. Following isotope injections in the anterior and lateral dorsal thalamic nuclei, labelled fibers form an arch in the white matter behind the cingulate sulcus and occupy the ventral sector of the cingulum bundle. The fibers from the anterior thalamic nucleus coursing in the cingulum bundle extended rostrally to the frontal cortex and caudally to area 23 and the retrosplenial cortex. In contrast, the fibers from lateral dorsal nucleus reached the retrosplenial cortex as well as the parahippocampal gyrus and presubiculum. Efferent fibers from the cingulate gyrus occupy a dorsolateral sector of the cingulum bundle. Those fibers from area 24 of the cingulate gyrus are directed to the premotor and prefrontal regions as well as area 23 and retrosplenial cortex. The fibers from area 23 extend rostrally to the prefrontal cortex and caudoventrally to the presubiculum and parahippocampal gyrus. Finally, an association component originates mainly from prefrontal cortex and posterior parietal region. These fibers occupy a more dorsal and lateral periphery in the cingulate white matter. Cingulum bundle fibers from the prefrontal cortex extend up to the retrosplenial cortex while those from the posterior parietal cortex extend caudally to the parahippocampal gyrus and presubiculum, and rostrally up to the prefrontal cortex.The Journal of Comparative Neurology 06/1984; 225(1):31-43. · 3.66 Impact Factor
Regionally Specific White Matter Disruptions of Fornix
and Cingulum in Schizophrenia
Muhammad Farid Abdul-Rahman1, Anqi Qiu1,2,3*, Kang Sim4,5
1Division of Bioengineering, National University of Singapore, Singapore, Singapore, 2Clinical Imaging Research Center, National University of Singapore, Singapore,
Singapore, 3Singapore Institute for Clinical Sciences, the Agency for Science, Technology and Research, Singapore, Singapore, 4Research Department, Institute of Mental
Health, Singapore, Singapore, 5Department of General Psychiatry, Institute of Mental Health, Singapore, Singapore
Limbic circuitry disruptions have been implicated in the psychopathology and cognitive deficits of schizophrenia, which
may involve white matter disruptions of the major tracts of the limbic system, including the fornix and the cingulum. Our
study aimed to investigate regionally specific abnormalities of the fornix and cingulum in schizophrenia using diffusion
tensor imaging (DTI). We determined the fractional anisotropy (FA), radial diffusivity (RD), and axial diffusivity (AD) profiles
along the fornix and cingulum tracts using a fibertracking technique and a brain mapping algorithm, the large deformation
diffeomorphic metric mapping (LDDMM), in the DTI scans of 33 patients with schizophrenia and 31 age-, gender-, and
handedness-matched healthy controls. We found that patients with schizophrenia showed reduction in FA and increase in
RD in bilateral fornix, and increase in RD in left anterior cingulum when compared to healthy controls. In addition, tract-
based analysis revealed specific loci of these white matter differences in schizophrenia, that is, FA reductions and AD and RD
increases occur in the region of the left fornix further from the hippocampus, FA reductions and RD increases occur in the
rostral portion of the left anterior cingulum, and RD and AD increases occur in the anterior segment of the left middle
cingulum. In patients with schizophrenia, decreased FA in the specific loci of the left fornix and increased AD in the right
cingulum adjoining the hippocampus correlated with greater severity of psychotic symptoms. These findings support
precise disruptions of limbic-cortical integrity in schizophrenia and disruption of these structural networks may contribute
towards the neural basis underlying the syndrome of schizophrenia and clinical symptomatology.
Citation: Abdul-Rahman MF, Qiu A, Sim K (2011) Regionally Specific White Matter Disruptions of Fornix and Cingulum in Schizophrenia. PLoS ONE 6(4): e18652.
Editor: Wang Zhan, University of California, San Francisco, United States of America
Received October 26, 2010; Accepted March 12, 2011; Published April 14, 2011
Copyright: ? 2011 Abdul-Rahman et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: The work reported here was supported by grants A*STAR SERC 082-101-0025, A*STAR SICS-09/1/1/001, a centre grant from the National Medical
Research Council (NMRC/CG/NUHS/2010), the Young Investigator Award at National University of Singapore (NUSYIA FY10 P07), and National University of
Singapore MOE AcRF Tier 1. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing Interests: The authors have declared that no competing interests exist.
* E-mail: firstname.lastname@example.org
Limbic circuitry disruptions have been implicated in the
psychopathology and cognitive deficits of schizophrenia [1,2,3].
The fornix and cingulum bundles are the most prominent white
matter fiber tracts within the limbic system. The fornix is a major
white matter bundle projection from the hippocampus to other
brain structures, including the mamillary bodies, thalamus, septal
region, nucleus accumbens, whereas the cingulum connects the
cingulate cortex with the prefrontal cortex, premotor regions,
cortical association areas in the parietal and occipital lobes,
parahippocampal cortex and the thalamus [4,5]. With regard to
the cingulate gyrus, earlier post-mortem studies have observed
altered neuronal arrangement , smaller pyramidal cells  and
reductions in the oligodendrocytes within this brain region .
Structural MRI studies have reported gray matter abnormalities in
the cingulate gyrus, especially in the anterior segment in
schizophrenia [9,10,11,12]. With respect to the fornix, previous
neuropathological studies have found no difference in fornix
volume , but increased fiber density of the left fornix in male
patients with schizophrenia . Looking at early onset cases,
Davies et al  found an increase in the cross sectional area of
the fornix in his study of subjects with schizophrenia. Taken
together, these data suggest that structural disruptions of the fornix
and cingulum bundles may occur in schizophrenia.
Diffusion tensor imaging (DTI) is a specific neuroimaging
technique that enables measure of the restricted water diffusion in
brain tissue. Fractional anisotropy (FA) derived from the diffusion
tensor may be influenced by myelination, orientation, coherence,
packing density and structural integrity of neural fiber tracts
[16,17]. Hence, reduction in FA often points to possible structural
abnormalities in neural fiber tracts. To better understand
biological processes behind changes in FA, radial diffusivity (RD)
and axial diffusivity (AD) derived from the diffusion tensor model
can be utilized to predict pathophysiological disruptions such as
demyelination [18,19] or axonal damage .
Voxel or region of interest  based assessments of the
cingulum and fornix bundles have been employed to study FA
alterations in schizophrenia [1,22,23,24,25], primarily pursuing
the hypothesis that schizophrenia would be associated with FA
differences in these two bundles. Reductions of FA in the cingulum
have been found in some [21,26,27] but not all DTI studies of
patients with schizophrenia [28,29]. Extant examinations of the
fornix using DTI have found reductions of FA in schizophrenia
[1,25,30] but Kendi et al  specifically investigated this limbic
bundle and did not find differences in FA in their study of early
PLoS ONE | www.plosone.org1 April 2011 | Volume 6 | Issue 4 | e18652
onset patients with schizophrenia. The observed discrepancies
between studies may be due to clinical differences in the
populations studied as well as methodological differences in
anatomical definitions of these two bundles. Misalignment of
anatomy in voxel-based analysis and segmentation errors of white
matter bundles in ROI-based analysis may exaggerate partial
volume effects. As consequences, they may increase random effects
on FA and thus decrease statistical power in detecting abnormal-
ities of the fornix and cingulum regions in schizophrenia. One
potential way to resolve this is to reliably delineate the core of these
fiber bundles and to localize FA abnormalities along the entire
white matter bundle.
To date, efforts to examine the white matter differences within
the limbic white matter bundles (cingulum and fornix) in a detailed
manner covering the entire tracts have been sparse. Recently,
Segal et al  manually traced the anterior and posterior
cingulate gyri, divided the anterior cingulate gyrus axially into six
equal segments, and the posterior cingulate gyrus into two
segments in their quest to understand the regional differences
within the cingulate gyrus. Unlike earlier approaches [23,24,25]
that have largely limited FA quantification to averaged values
within the ROIs of these fiber tracts, we mapped FA, RD, and AD
values along the entire fiber bundles of the fornix and cingulum
using a tractographic approach  and with geometric
representation of these bundles generated using large deformation
diffeomorphic metric mapping (LDDMM)  in this study. In
particular, we employed fibertracking on an averaged DTI over
the subjects in our study such that the fornix and cingulum were
delineated only in the region that is most overlapped across our
samples. This potentially reduces partial volume effects on
statistical testing of the diffusion measurements. Additionally, our
method facilitated the detection of FA, RD, and AD differences
with respect to the structural geometry of each fiber bundle, which
can potentially clarify regionally specific white matter abnormal-
ities within the fornix and cingulum in schizophrenia.
Materials and Methods
Thirty three right-handed patients with schizophrenia and
thirty one age-, gender- and handedness-matched healthy controls
were recruited from the Institute of Mental Health, Singapore and
the community respectively for this study. The study was approved
by the Institutional Review Boards of the Institute of Mental
Health, Singapore, as well as that of the National Neuroscience
Institute, Singapore. All subjects gave their written informed
consent following a complete description of the study. Socio-
demographic and clinical information for the two groups of
subjects are given in Table 1.
All diagnoses were made by a psychiatrist using information
obtained from the clinical history, mental status examination,
existing medical records, interviews with significant others as well
as the administration of the Structured Clinical Interview for
DSM-IV disorders-Patient Version (SCID-P) . The patients
were maintained on a stable dose of antipsychotic medications for
at least two weeks prior to recruitment and did not have their
medications withdrawn for the purpose of the study. Twenty five
patients received second generation antipsychotics, seven patients
were prescribed first generation antipsychotics and one patient was
on a combination of first and second generation antipsychotics.
The mean (SD) antipsychotic dose was 200.15 (153.73) daily
chlorpromazine equivalents in milligrams. No subject met DSM-
IV criteria for alcohol within the three months prior to the scan or
other lifetime substance abuse. The Positive and Negative
Syndrome Scale (PANSS)  was used to assess the nature and
severity of psychopathology, while the Global Assessment of
Functioning Scale  was used to assess the level of psychosocial
functioning. Both scales were administered by a psychiatrist to all
the participants. The healthy controls were screened using the
SCID-NP  to be free of any Axis I psychiatric disorder. None
of the subjects had a history of major neurological, medical
illnesses, substance abuse or psychotropic medication use.
2.2 DTI acquisition and preprocessing
Single-shot echo-planar DTI (TR=3725 ms; TE=56 ms; flip
angle=90u) was acquired using a 3-Tesla whole body Philips
scanner with a SENSE head coil. 42 axial slices with 3.0 mm
thickness were acquired parallel to the anterior–posterior com-
missure line; the imaging matrix was 1126109 with a field of view
of 230 mm6230 mm, which was zero-filled to 2566256. 15
diffusion weighted images (DWIs) with b=800 sec/mm2and 1
baseline with b=0 sec/mm2were obtained.
Within each subject, DWIs were first corrected for motion and
eddy current distortions using affine transformation to the image
without diffusion weighting. These DWIs were then registered to
Mori’s single-subject DTI atlas (resolution: 16161 mm3, http://
www.mristudio.org)  via affine transformation between the
images without diffusion weighting. To align subjects’ diffusion
tensor images to Mori’s atlas, we employed the LDDMM image
mapping to simultaneously align the subject’s FA map and the
image without diffusion weighting to those of Mori’s single-subject
Table 1. Demographic and clinical characteristics of the sample.
(N=33)Test statistic P value
Age (SD), years35.4 (8.82) 39.4 (8.82)t62=21.844 0.0744
Sex (% Male) 77.4% (24/31)76.0% (25/33)
Years of Education (SD), years13.9 (2.50) 11.6 (2.40)t62=3.7000.0005
Mean Illness Duration (SD), years- 12.7 (8.99)--
PANSS positive symptom subscale- 10.1(3.28)--
PANSS negative symptom subscale- 8.7(2.68)--
PANSS general psychopathology subscale- 19.9 (4.17)--
GAF scores- 53.6 (17.5)--
Abbreviations: CON, control subjects; SCZ, patients with schizophrenia; SD, standard deviation.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org2 April 2011 | Volume 6 | Issue 4 | e18652
DTI atlas . We followed the same mapping procedure and
LDDMM parameter setting as those given in . The accuracy
of this nonlinear registration approach has been extensively
validated and reported elsewhere . For each individual
subject, the tensor, FA, RD, and AD maps were recomputed
using the DWIs aligned to Mori’s DTI atlas via the diffeomorphic
transformation. In addition, the mean DWIs were also constructed
by averaging the corresponding DWIs across healthy controls. The
tensor and FA map were computed from the mean DWIs in
Mori’s atlas space to represent the white matter anatomy in this
2.3 Fornix and Cingulum Fiber Tracking
The fornix and cingulum bundles in each hemisphere were
reconstructed from the mean DTI using the Fiber Assignment
Continuous Tracking (FACT) method  with a FA threshold of
0.15 and an angle threshold of 50u. The FACT tracking was first
performed from all pixels inside the brain. Then, each bundle of
interest was extracted using a multi-ROI approach, that is, all
tracts penetrating these ROIs were assigned to this bundle for
further analysis. The protocol that defines these ROIs in the mean
DTI color map, which is in Mori’s DTI atlas space, for the fornix
and cingulum bundles is elaborated below. The ROI drawing was
repeatable given their coordinates in the Mori’s DTI atlas space.
Two ROIs were manually drawn on the color map.
The first ROI was placed on the most anterior coronal plane
where the body of fornix is visible on the color map (Figure 1(b),
coronal slice 150). Then, the second ROI was drawn on the axial
plane that includes the tail of the hippocampus (Figure 1(c), axial
slice 108). After the fornix bundle was extracted, the average
values of FA, AD, and RD were calculated for the entire bundle,
separately for each hemisphere.
The cingulum bundle can be divided into two
segments, namely, the upper part along the main cingulate gyrus
(CGC: cingulum in the cingulate gyrus) and the lower segment
along the ventral side of the hippocampus (CGH: cingulum
adjoining the hippocampus) . Two ROIs were used to
extract each segment of the cingulum bundle. For CGH, it runs
along the ventral aspect of the hippocampus. The first ROI was
defined on the axial plane where the cingulum bundle begins to
turn anteriorly above the splenium (Figure 2(c), axial slice 126)
and the second ROI was chosen on the coronal plane anterior to
the pons (Figure 2(d), coronal slice 125). For CGC, the first
ROI was defined on the axial plane where CGH ends
(Figure 2(c), axial slice 126) and the second ROI was defined
on the coronal slice where is the most anterior to the corpus
callosum (Figure 2(g), coronal slice 183). CGC was further
divided into three portions (anterior, middle, and posterior)
based on the cortical projection of the corpus callosum. The two
(Figure 2(e, f), coronal slices 127 and 162) where the fiber
orientation of the corpus callosum was changed (Figure 2(b)).
After the four bundles were extracted, the average values of FA,
RD and AD were calculated for each bundle, separately for each
extracted using the FACT fibertracking approach on the
averaged DTI. The coordinates of the multiple ROIs in Mori’s
DTI atlas where each bundle passes through were recorded and
easily reproduced using the fibertracking method. However, since
the tracking was performed only on the averaged DTI,
misalignment of the DTI data across subjects could affect the
result of the fiber tracking, especially for small structures like the
fornix. To ensure that the fornix of each subject overlapped with
the ROI of the fornix extracted from the averaged DTI using the
fiber tracking method, we manually segmented the fornix of five
subjects (see manual segmentation in Figure 3) and transformed
the fornix mask to the atlas. The volume overlap ratio between the
manual segmentation volume and the automatically extracted
volume in Mori’s atlas was computed. Its mean and standard
deviation are 87.5% and 2.6%, indicating high reliability in the
procedure of the fornix delineation within this study.
Mean Curve Generation.
was computed using the LDDMM curve mapping [34,41] based
on all the fiber tracts within this bundle. The mean curve was then
parameterized at 30 points along its arc length and was used as the
geometric representation of the bundle. The FA, RD and AD
distributions along each bundle was characterized as a function of
the arc length, and obtained for each subject. This allows the
detection of local abnormalities at specific locations along each
bundle of interest.
cingulum bundles were
For each bundle, a mean curve
2.4 Statistical Analysis
Demographic variables between patients with schizophrenia
and control subjects were compared using two sample Student’s
t-test for continuous variables and chi-square test for categorical
Linear regression with the main effects of diagnostic group
(control vs schizophrenia) and covariates of age, years of
education, and illness duration were performed on the mean
FA, RD and AD values in each bundle to elucidate abnormalities
Figure 1. Delineation of the fornix bundle (Fx). Panel (a) illustrates the 3D view of the brain, where the fornix, hippocampus, corpus callosum
are colored in yellow, pink, and green, respectively. Panels (b, c) show the coronal and axial slices of the mean DTI color map, where the ROIs are
defined for delineating the fornix.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org3April 2011 | Volume 6 | Issue 4 | e18652
To detect local differences for values of FA, RD and AD
between the two diagnostic groups, the respective values at each
point on the mean curve of each bundle was first smoothed by
averaging the values of its closest five neighboring points along the
curve. Linear regression with the main effects of diagnostic group
(control versus schizophrenia) and covariates of age, years of
education, and illness duration were then used to examine local
differences in FA, RD and AD between the two diagnostic groups.
The statistical results were visualized at an overall significance
level of 0.05 after Bonferroni correction.
Correlation with Clinical Measures
Partial correlation analysis was performed to investigate the
relationship between the DTI measures within each ROI and
psychopathological measures assessed by PANSS in patients
with schizophrenia when controlling for age, years of education,
and illness duration. To detect the specific location along each
bundle where the DTI measure is associated with the
psychopathology, the same partial correlation analysis was
performed on every point of each bundle. The statistical results
were visualized at an overall significance level of 0.05 after
Figure 2. Delineation of the cingulum bundles. They include the segment adjoining the hippocampus (CGH) and the main cingulum (CGC).
Panel (a) illustrates the 3D view of the brain, where the cingulum, hippocampus, and corpus callosum are colored in yellow, pink, and green. Panels
(b–g) respectively show the sagittal, axial, and coronal slices of the average DTI color map, where the ROIs are defined for delineating CGH and CGC.
CGH is defined by the ROIs given in panels (c, d) whereas CGC is defined by the ROIs given in panels (c, g). CGC is further divided into three partitions
using the ROIs in panels (b, e, f).
Figure 3. Fornix manual segmentation. Panel (a) shows the fornix mask extracted from the averaged DTI using the fiber tracking method. Panels
(b–f) show the manual masks of the fornix labeled on the DTI of five subjects.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org4 April 2011 | Volume 6 | Issue 4 | e18652
Demographic and Clinical Characteristics
Compared to healthy controls, patients with schizophrenia had
fewer years of education (p=0.0005). There was no statistically
significant difference in age (p=0.0744) and sex (p=0.8754)
between the two groups. The average duration of illness in patients
with schizophrenia was 12.7 years.
Figure 4 and 5 illustrate the box plots of the mean FA, mean
RD and mean AD values in the fornix and cingulum bundles,
respectively. After controlling for age, years of education, and
illness duration, patients with schizophrenia had statistically
significant FA reductions bilaterally in the fornix bundles (left:
p,0.0005; right: p=0.0492) compared to healthy controls. This
was accompanied by increase in RD (left: p,0.0005; right:
p=0.0061) but not increase in AD of the fornix bundle (left:
p=0.9321; right: p=0.8848). For the CGC bundle, patients with
schizophrenia showed statistically significant RD increase in the
left anterior CGC (p=0.0041) with less significant decrease in FA
bilaterally (left: p=0.0563; right: p=0.0487). No other group
difference was found using ROI-based analysis.
Figure 6 illustrates the fornix and cingulum bundles and their
mean tracts in the diffusion tensor images averaged over all the
control subjects in this study. The FA, RD, and AD distributions
along these mean curves are plotted in Figure 7 with solid and
dashed lines representing the averaged values in healthy controls
and patients with schizophrenia respectively. Shaded rectangles
correspond to regions with p,0.05 after Bonferroni correction.
Compared to control subjects, patients with schizophrenia
showed reductions in FA along the specific regions of the left
fornix (Figure 7(a)). It was also found that the RD and AD
increased along middle region and superior region of the left fornix
(Figure 7(f, k)). In the same location of the right fornix, patients
with schizophrenia also showed a trend of the FA reduction and
RD and AD increase although not at a significant level when
compared with healthy controls (Figure 7(p, u, z)).
For the CGC bundle, the rostral region of the left anterior CGC
showed lower FA and higher RD in patients with schizophrenia
than in control subjects but no significant difference in AD
(Figure 7(e,j,o)). In addition, patients with schizophrenia showed
significant increases in both RD and AD but no significant FA
reduction in the anterior segment of the left middle CGC
(Figure 7(d,i,n)). No significant FA reduction or RD or AD
increases were found in the left posterior CGC and CGH as well
as the right CGC bundles.
Relations with Clinical Measures
Even though the ROI-based analysis did not reveal significant
correlations of the PANSS scores with any DTI measures, the
tract-based analysis found significant negative correlations of the
FA values in the loci of left fornix near to the hippocampus with
PANSS total score (Figure 8(a)), PANSS positive symptom score
(Figure 8(b)), and PANSS general psychopathology score
(Figure 8(c)), indicating that a lower FA value correlated with
greater psychopathology as shown by higher PANSS total, positive
symptom, and general psychopathology scores. Additionally,
significant positive correlations of the AD values with PANSS
total score (Figure 8(d)) and PANSS positive symptom score
(Figure 8(e)) were found at the loci of the right CGH near to the
splenium, indicating that a larger AD value correlated with more
severe symptomatology as shown by higher PANSS total and
positive symptom scores.
This is the first study on schizophrenia to map regionally
specific FA, RD, and AD differences along the entire fornix and
cingulum in a detailed manner. In the traditional ROI-based
analysis, we found significant reductions in FA and increases in
RD in bilateral fornix, and RD increase in the left anterior
cingulum in patients with schizophrenia compared to controls.
Tract-based analysis further revealed specific loci of these white
matter differences in schizophrenia in that FA reductions and RD
increases occur in the region of the left fornix more distal from the
hippocampus and in the rostral portion of the left anterior
cingulum. In addition, the white matter properties at specific loci
of the left fornix and the right cingulum were associated with
psychopathology in patients with schizophrenia. These findings
support precise disruptions of limbic-cortical integrity in schizo-
Our finding of the mean FA reduction in the left fornix of
patients with schizophrenia using tract-based analysis was
consistent with findings of earlier studies where tract-based
analysis were used [1,25] but in contrast to studies using voxel-
based analysis [42,43]. Kuroki et al  specifically examined
fornix abnormalities using a ROI approach and found reduction
Figure 4. ROI-based analysis of the fornix DTI measures. Panels (a, b, c) respectively show the box plots of the FA, RD and AD values averaged
over left and right fornix regions of interest. Black and blue boxes represent the values in the control and schizophrenia groups, respectively. Two,
and three red asterisks respectively denote the p-value less than 0.05, 0.01.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org5 April 2011 | Volume 6 | Issue 4 | e18652
of FA within the fornix in schizophrenia, and together with data
from tract-based studies, suggest that more precise definition and
evaluation of the fornix are needed to better appreciate underlying
white matter differences. In this study, more detailed examination
of FA differences along the fornix added into the literature that FA
reduction in schizophrenia occurred in the crux of the left fornix
towards the midline where it merges with the crux of the right
fornix and forms the body of fornix, which may be due to
underlying demyelination, axonal damage or a combination of
both processes as indicated by increases in RD and AD. Kumar
et al.  speculated that the coincident increase of RD and AD is
suggestive of reduction in axonal density because the consequent
increase in axonal space results in less convergence of neural fiber
tracts. Taken together, these evidences reinforced the theory of
Figure 5. ROI-based analysis of the cingulum DTI measures. Panels (a, c, e) and (b, d, f) respectively show the box plots of the FA, RD and AD
values averaged over left and right cingulum regions of interest. Black and blue boxes represent the values in the control and schizophrenia groups,
respectively. One, two and three red asterisks respectively denote the p-value less than 0.1, 0.05 and 0.01. Abbreviations: CGH, cingulum adjoining
the hippocampus; aCGC, anterior cingulum; mCGC, middle cingulum; pCGC, posterior cingulum.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org6 April 2011 | Volume 6 | Issue 4 | e18652
demyelination or caliber reduction in the crux of the left fornix in
schizophrenia. While we observed significant reduction in the
mean FA in the right fornix based on ROI analysis, there is only a
trend of significance for FA decrease, RD and AD increase along
the right fornix bundle using tract-based analysis.
Based on tract-based analysis, mean FA reduction in the left
anterior CGC in schizophrenia is consistent with previous findings
implicating the anterior cingulate region in schizophrenia [3,21,26].
This significant FA reduction in the rostral region of left anterior
cingulum added to our understanding of cingulum pathology in
Figure 6. The fornix and cingulum fiber tracts (yellow lines) and their mean curves (blue line). Panels (a, b) respectively illustrate the left
and right fornix, where ‘‘H’’ and ‘‘F’’ indicate the orientation of the fornix bundle from the hippocampus to the thalamus. Panels (c, d) respectively
illustrate the left and right cingulum in the hippocampus (CGH), where ‘‘S’’ and ‘‘O’’ indicate the orientation of CGH from the splenium of the corpus
callosum to the hippocampus. Panels (e–j) show the left and right cingulum in the cingulate (CGC) from the anterior, middle, to the posterior
segments, where ‘‘A’’, ‘‘M’’, and ‘‘P’’ indicate the orientation of CGC from the anterior to the posterior.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org7 April 2011 | Volume 6 | Issue 4 | e18652
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org8 April 2011 | Volume 6 | Issue 4 | e18652
terms of information about the precise location of white matter
integrity changes in schizophrenia [3,46]. Concurrently, the rostral
region of the left anterior CGC also showed increased RD in
schizophrenia, which together with FA reduction suggests that
demyelination may underlie the white matter aberrations.
Regionally specific increase in RD was also found in the rostral
region of the left middle CGC. There was a simultaneous increase
in AD in this region, which was not picked up by the ROI-based
analysis of the entire left middle CGC bundle. These results point
to abnormalities in the bundle, which may be due to altered
cytoarchitecture leading to increased CSF in this specific region.
Compared to the anterior and posterior CGC, the middle CGC
showed relatively higher FA and AD and lower RD. The relatively
higher mean FA in the middle CGC compared to the anterior and
posterior CGC may be related to the composition of long and
short neural fibers and the changes along their connections with
other brain regions . It is likely that the mean curve
representing the middle cingulum begins in a region with less
fiber branches and hence, maintains higher FA before descending
to both the anterior and posterior regions with more fiber
branches to the other brain structures. It also results in AD
decreasing and RD increasing simultaneously in the anterior and
posterior CGC when compared to the middle CGC.
In our study, the white matter differences in the cingulum
within schizophrenia were mainly limited to the anterior cingulum
but not posterior cingulum, which indicates disruption of
connections between the anterior cingulate and the prefrontal
cortex which is germane to observed deficits of executive
functioning in schizophrenia. Takei et al  found that white
matter disruption in the cingulum was associated with increased
reaction time in the Stroop task and Manoach et al  noted
reductions in FA of the anterior cingulate gyrus to be associated
with longer saccadic latencies in schizophrenia, underscoring the
significance of the cingulum integrity in visuospatial attention and
executing functioning in schizophrenia. In addition, the cingulum
white matter differences in our study were more marked on the left
cingulum, which was in line with most of earlier studies
[2,26,49,50]. This may indicate laterality effects which may be
Figure 8. Partial correlation analysis between DTI measures and clinical scores in left fornix and right CGH. Panels (a–c) respectively
show the correlation profile of FA values along the left fornix with PANSS total score, PANSS positive symptom score and PANSS general
psychopathology score. Panels (d–e) show the correlation profile of AD values along the right CGH with PANSS total score and PANSS positive
symptom score. The anatomical landmarks of each tract are given in Figure 6. Regions colored by gray are the anatomical locations with significant
group difference at p,0.05 after Bonferroni correction. The range of p-values is indicated on each panel.
Figure 7. Tract-based analysis of DTI measures along fornix and cingulum. The top three rows respectively show FA, RD and AD profiles for
left hemisphere, while the bottom rows show those for the right hemisphere. The panels from left to right respectively illustrate the profiles for
regions of the fornix (Fx), cingulum adjoining the hippocampus (CGH), posterior cingulum (pCGC), middle cingulum (mCGC), and anterior cingulum
tracts (aCGC). On each panel, solid line denotes the profile averaged over healthy controls, while dashed line denotes the profile averaged over
patients with schizophrenia. The anatomical landmarks of each tract are given in Figure 6. Regions colored by gray are the anatomical locations with
significant group difference at p,0.05 after Bonferroni correction. The range of p-values is indicated on each panel.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org9 April 2011 | Volume 6 | Issue 4 | e18652
related to predominant right handedness of our study cohort and/
or accentuation of left greater than right asymmetry involving
white matter FA differences .
The negative correlations between FA along specific region of the
left fornix and PANSS total, positive and general psychopathology
scores are in contrast to the previous finding of absence of correlation
between mean FA of fornix, delineated with the ROI-based
approach, and psychotic symptomatology . We did not find
correlation between FAalong thecingulumandPANSSscoreswhich
is in agreement with the findings from several [52,53,54] but not all
earlier studies . In view of the scarce findings, we admit that it is
possible that severity of symptoms is tied to a pattern of white matter
abnormalities that occurat specific locialong this fiber bundle instead
of uniformly at one region across all subjects. At the same time,
severity of schizophrenia symptoms may also be modulated by other
regions as suggested by Mitelman et al . Taken together, these
findings suggest that the specific disruptions of white matter integrity
along relevant neural tracts such as fornix and cingulum may exert a
dominant effect, which could underlie psychotic psychopathology in
studies suggested the hippocampus as the brain origin of schizophre-
nia [57,58]. Earlier neuroimaging studies revealed no cortical
thinning  but anterior hippocampal deformity  in unaffected
schizophrenic siblings, suggesting the hippocampal anatomy may
confer vulnerability to the illness. Qiu et al.  examined the
hippocampal morphology and its relation with the cortical thickness
in the same cohort as that in this study and suggested the
hippocampal-cortical disruptions in schizophrenia. Taken together
with our results, the relation of schizophrenia diagnostic scores with
the white matter disruption in the fornix near the hippocampus may
suggest that brain tissue damage may be propagated due to the loss of
white matter connections as schizophrenia symptoms become more
severe. We notice that the correlation of FA with schizophrenia
symptoms in the fornix is overlapped with the location of FA
abnormalities in schizophrenia (Figure 7(a)) but not exactly the same.
This may be due to the low degree of freedom since this correlation
analysis was only performed in patients with schizophrenia.
There were several methodological strengths and limitations in
this study. In terms of strengths, first, we employed the
tractography on the averaged DTI such that the fornix and
cingulum were delineated only in the subjects’ overlapped region.
Thus, it potentially reduced partial volume effects at the boundary
of each structure on DTI measures. Second, we specifically
examined the limbic white matter bundles using the DTI
tractographic approach in addition to ROI-based analysis. This
tractographic approach has higher specificity and sensitivity
compared with conventional ROI or VBM approaches  in
both the detection of abnormalities and the correlation of DTI
measures with clinical scores. The extraction of the fornix and
cingulum bundles is replicable based on Mori’s DTI atlas and the
delineation protocol as described above. Third, the combined
investigations of white matter disruptions along both the fornix
and the cingulum allows a more holistic understanding of the FA,
RD, and AD differences within these limbic pathways traversing
cortical and subcortical structures. Third, we analyzed the
variation in FA, RD, and AD as a function of geometric position
along the white matter bundle. Compared to the voxel-based
analysis across the whole brain, our method is more specific and
maps FA, RD, and AD differences along the entire fornix and
cingulum. Our findings may provide guidance of the spatial
location of resection in postmortem studies for understanding
neuropathology of the fornix and cingulum in schizophrenia. In
terms of limitations, our DTI data were not acquired in a cardiac
gated fashion, which may cause motion induced by natural cardiac
movement. As introduced in , the imaging acquisition could
be further improved for better characterizing the small structures,
such as the fornix. Furthermore, most of the patients with
schizophrenia included in this study had been prescribed
antipsychotic medications. However, earlier studies had not found
medication related effects on FA [52,62] and we did not find
significant correlations between the antipsychotic dose and
regional FA of these tracts. Future studies may want to better
detail the disruptions of white matter integrity of fornix and
cingulum in drug naı ¨ve individuals over illness course in order to
tease out effects of psychotropic medications on the white matter
In summary, we employed a tractographic approach to examine
region specific FA, RD, and AD differences in the fornix and
cingulum bundles. Our results suggested that regionally-specific
white matter differences occurred at left fornix and the left anterior
and middle cingulum in schizophrenia, which may be due to
underlying demyelination of these fiber bundle regions, resulting
in the disruption of limbic and cortical connectivities in
schizophrenia. Furthermore, decreased FA at the left fornix and
increased AD at right cingulum correlated with greater severity of
psychotic symptoms in patients with schizophrenia, suggesting that
disruption of white matter integrity may contribute towards the
neural basis of clinical symptomatology. Considering the role of
the hippocampus and cingulate gyrus in schizophrenia neurobi-
ology, further investigation of region specific brain gray and white
matter changes over time is needed to better understand the
interactional nature of the relationship between these limbic
structures and other related cortical and subcortical brain
structures in the context of illness onset, progression and response
Conceived and designed the experiments: MA AQ KS. Performed the
experiments: MA AQ KS. Analyzed the data: MA AQ KS. Contributed
reagents/materials/analysis tools: MA AQ KS. Wrote the paper: MA AQ
KS. Collected the samples: KS. Conducted the data analysis and paper
writing: MA KS AQ.
1. Takei K, Yamasue H, Abe O, Yamada H, Inoue H, et al. (2008) Disrupted
integrity of the fornix is associated with impaired memory organization in
schizophrenia. Schizophrenia Research 103: 52–61.
2. Nestor PG, Kubicki M, Spencer KM, Niznikiewicz M, McCarley RW, et al.
(2007) Attentional networks and cingulum bundle in chronic schizophrenia.
Schizophr Res 90: 308–315.
3. Fujiwara H, Namiki C, Hirao K, Miyata J, Shimizu M, et al. (2007) Anterior
and posterior cingulum abnormalities and their association with psychopathol-
ogy in schizophrenia: A diffusion tensor imaging study. Schizophrenia Research
4. Mufson EJ, Pandya DN (1984) Some observations on the course and composition
of the cingulum bundle in the rhesus monkey. J Comp Neurol 225: 31–43.
5. Mori S, Oishi K, Jiang H, Jiang L, Li X, et al. (2008) Stereotaxic white matter atlas
basedondiffusiontensorimaginginanICBM template. Neuroimage 40:570–582.
6. Benes FM, Majocha R, Bird ED, Marotta CA (1987) Increased vertical axon
numbers in cingulate cortex of schizophrenics. Archives of General Psychiatry
7. Bouras C, Ko ¨vari E, Hof PR, Riederer BM, Giannakopoulos P (2001) Anterior
cingulate cortex pathology in schizophrenia and bipolar disorder. Acta
Neuropathologica 102: 373–379.
8. Stark AK, Uylings HBM, Sanz-Arigita E, Pakkenberg B (2004) Glial Cell Loss in
the Anterior Cingulate Cortex, a Subregion of the Prefrontal Cortex, in Subjects
With Schizophrenia. Am J Psychiatry 161: 882–888.
9. Baiano M, David A, Versace A, Churchill R, Balestrieri M, et al. (2007) Anterior
cingulate volumes in schizophrenia: a systematic review and a meta-analysis of
MRI studies. Schizophr Res 93: 1–12.
10. Qiu A, Younes L, Wang L, Ratnanather JT, Gillepsie SK, et al. (2007)
Combining anatomical manifold information via diffeomorphic metric map-
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org10April 2011 | Volume 6 | Issue 4 | e18652
pings for studying cortical thinning of the cingulate gyrus in schizophrenia.
Neuroimage 37: 821–833.
11. Wang L, Hosakere M, Trein JC, Miller A, Ratnanather JT, et al. (2007)
Abnormalities of cingulate gyrus neuroanatomy in schizophrenia. Schizophr Res
12. Fornito A, Yucel M, Wood SJ, Adamson C, Velakoulis D, et al. (2008) Surface-
based morphometry of the anterior cingulate cortex in first episode
schizophrenia. Hum Brain Mapp 29: 478–489.
13. Brisch R, Bernstein H-G, Stauch R, Dobrowolny H, Krell D, et al. (2008) The
volumes of the fornix in schizophrenia and affective disorders: A post-mortem
study. Psychiatry Research: Neuroimaging 164: 265–273.
14. Chance SA, Highley JR, Esiri MM, Crow TJ (1999) Fiber Content of the Fornix
in Schizophrenia: Lack of Evidence for a Primary Limbic Encephalopathy.
Am J Psychiatry 156: 1720–1724.
15. Davies DC, Wardell AM, Woolsey R, James AC (2001) Enlargement of the
fornix in early-onset schizophrenia: a quantitative MRI study. Neurosci Lett
16. Basser PJ (1995) Inferring microstructural features and the physiological state of
tissues from diffusion-weighted images. NMR in Biomedicine 8: 333–344.
17. Pierpaoli C, Jezzard P, Basser PJ, Barnett A, Di Chiro G (1996) Diffusion tensor
MR imaging of the human brain. Radiology 201: 637–648.
18. Loy DN, Kim JH, Xie M, Schmidt RE, Trinkaus K, et al. (2007) Diffusion
Tensor Imaging Predicts Hyperacute Spinal Cord Injury Severity. Journal of
Neurotrauma 24: 979–990.
19. Song S-K, Sun S-W, Ramsbottom MJ, Chang C, Russell J, et al. (2002)
Dysmyelination Revealed through MRI as Increased Radial (but Unchanged
Axial) Diffusion of Water. NeuroImage 17: 1429–1436.
20. Song S-K, Sun S-W, Ju W-K, Lin S-J, Cross AH, et al. (2003) Diffusion tensor
imaging detects and differentiates axon and myelin degeneration in mouse optic
nerve after retinal ischemia. NeuroImage 20: 1714–1722.
21. Takei K, Yamasue H, Abe O, Yamada H, Inoue H, et al. (2009) Structural
disruption of the dorsal cingulum bundle is associated with impaired Stroop
performance in patients with schizophrenia. Schizophrenia Research 114:
22. Kanaan RA, Kim JS, Kaufmann WE, Pearlson GD, Barker GJ, et al. (2005)
Diffusion tensor imaging in schizophrenia. Biol Psychiatry 58: 921–929.
23. Peters BD, de Haan L, Dekker N, Blaas J, Becker HE, et al. (2008) White matter
fibertracking in first-episode schizophrenia, schizoaffective patients and subjects
at ultra-high risk of psychosis. Neuropsychobiology 58: 19–28.
24. Rosenberger G, Kubicki M, Nestor PG, Connor E, Bushell GB, et al. (2008)
Age-related deficits in fronto-temporal connections in schizophrenia: a diffusion
tensor imaging study. Schizophr Res 102: 181–188.
25. Fitzsimmons J, Kubicki M, Smith K, Bushell G, Estepar RS, et al. (2009)
Diffusion tractography of the fornix in schizophrenia. Schizophr Res 107:
26. Wang F, Sun Z, Cui L, Du X, Wang X, et al. (2004) Anterior Cingulum
Abnormalities in Male Patients with Schizophrenia Determined Through
Diffusion Tensor Imaging. American Journal of Psychiatry 161: 573–575.
27. Hoptman MJ, Nierenberg J, Bertisch HC, Catalano D, Ardekani BA, et al.
(2008) A DTI study of white matter microstructure in individuals at high genetic
risk for schizophrenia. Schizophrenia Research 106: 115–124.
28. Mun ˜oz Maniega S, Lymer GKS, Bastin ME, Marjoram D, Job DE, et al. (2008)
A diffusion tensor MRI study of white matter integrity in subjects at high genetic
risk of schizophrenia. Schizophrenia Research 106: 132–139.
29. Narr KL, Hageman N, Woods RP, Hamilton LS, Clark K, et al. (2009) Mean
diffusivity: A biomarker for CSF-related disease and genetic liability effects in
schizophrenia. Psychiatry Research - Neuroimaging 171: 20–32.
30. Davenport ND, Karatekin C, White T, Lim KO (2010) Differential fractional
anisotropy abnormalities in adolescents with ADHD or schizophrenia.
Psychiatry Research - Neuroimaging 181: 193–198.
31. Kendi M, Karagulle Kendi AT, Lehericy S, Ducros M, Lim KO, et al. (2008)
Structural and diffusion tensor imaging of the fornix in childhood- and
adolescent-onset Schizophrenia. Journal of the American Academy of Child and
Adolescent Psychiatry 47: 826–832.
32. Segal D, Haznedar MM, Hazlett EA, Entis JJ, Newmark RE, et al. (2010)
Diffusion tensor anisotropy in the cingulate gyrus in schizophrenia. NeuroImage.
33. Mori SCBJ, Chacko VP, van Zijl PC (1999) Three-dimensional tracking of
axonal projections in the brain by magnetic resonance imaging. Annals of
neurology 45: 265–269.
34. Glaune `s J, Qiu A, Miller MI, Younes L (2008) Large deformation diffeomorphic
metric curve mapping. International Journal of Computer Vision 80: 317–336.
35. First MB, Frances AJ, Pincus HA, Vettorello N, Davis WW (1994) DSM-IV in
progress. Changes in substance-related, schizophrenic, and other primarily adult
disorders. Hosp Community Psychiatry 45: 18–20.
36. Kay SR, Fiszbein A, Opler LA (1987) The positive and negative syndrome scale
(PANSS) for schizophrenia. Schizophr Bull 13: 261–276.
37. Gaffan D (1994) Role of the amygdala in picture discrimination learning with
24-h intertrial intervals. Experimental Brain Research 99: 423–430.
38. First MB, Spitzer RL, Gibbon M, Williams JBW (2002) Structured Clinical
Interview for DSM-IV Axis I Disorders-Non-patient version (SCID-NP).
Washington, DC: American Psychiatric Press.
39. Ceritoglu C, Oishi K, Li X, Chou MC, Younes L, et al. (2009) Multi-contrast
large deformation diffeomorphic metric mapping for diffusion tensor imaging.
Neuroimage 47: 618–627.
40. Wakana S, Caprihan A, Panzenboeck MM, Fallon JH, Perry M, et al. (2007)
Reproducibility of quantitative tractography methods applied to cerebral white
matter. Neuroimage 36: 630–644.
41. Qiu A, Miller MI (2007) Cortical hemisphere registration via large deformation
diffeomorphic metric curve mapping. Med Image Comput Comput Assist Interv
Int Conf Med Image Comput Comput Assist Interv 10: 186–193.
42. Agartz IAJL, Skare S (2001) Abnormal brain white matter in schizophrenia: a
diffusion tensor imaging study. Neuroreport 12: 2251–2254.
43. Ardekani BA, Nierenberg J, Hoptman MJ, Javitt DC, Lim KO (2003) MRI
study of white matter diffusion anisotropy in schizophrenia. Neuroreport 14:
44. Kuroki N, Kubicki M, Neston PG, Salisbury DF, Park HJ, et al. (2006) Fornix
Integrity and Hippocampal Volume in Male Schizophrenic Patients. Biological
Psychiatry 60: 22–31.
45. Kumar R, Macey PM, Woo MA, Harper RM (2010) Rostral brain axonal injury
in congenital central hypoventilation syndrome. Journal of Neuroscience
Research 88: 2146–2154.
46. Kubicki M, McCarley R, Westin CF, Park HJ, Maier S, et al. (2007) A review of
diffusion tensor imaging studies in schizophrenia. J Psychiatr Res 41: 15–30.
47. Schmahmann JD, Pandya DN, Wang R, Dai G, D’Arceuil HE, et al. (2007)
Association fibre pathways of the brain: parallel observations from diffusion
spectrum imaging and autoradiography. Brain 130: 630–653.
48. Manoach DS, Ketwaroo GA, Polli FE, Thakkar KN, Barton JJS, et al. (2007)
Reduced microstructural integrity of the white matter underlying anterior
cingulate cortex is associated with increased saccadic latency in schizophrenia.
NeuroImage 37: 599–610.
49. Kumra S, Ashtari M, Cervellione KL, Henderson I, Kester H, et al. (2005)
White matter abnormalities in early-onset schizophrenia: A voxel-based diffusion
tensor imaging study. Journal of the American Academy of Child and
Adolescent Psychiatry 44: 934–941.
50. Hao Y, Yan Q, Liu H, Xu L, Xue Z, et al. (2009) Schizophrenia patients and
their healthy siblings share disruption of white matter integrity in the left
prefrontal cortex and the hippocampus but not the anterior cingulate cortex.
Schizophrenia Research 114: 128–135.
51. Luck D, Malla AK, Joober R, Lepage M (2010) Disrupted integrity of the fornix
in first-episode schizophrenia. Schizophrenia Research 119: 61–64.
52. Foong J, Maier M, Clark CA, Barker GJ, Miller DH, et al. (2000)
Neuropathological abnormalities of the corpus callosum in schizophrenia: A
diffusion tensor imaging study. Journal of Neurology Neurosurgery and
Psychiatry 68: 242–244.
53. Minami T, Nobuhara K, Okugawa G, Takase K, Yoshida T, et al. (2003)
Diffusion Tensor Magnetic Resonance Imaging of Disruption of Regional White
Matter in Schizophrenia. Neuropsychobiology 47: 141–145.
54. Skelly LR, Calhoun V, Meda SA, Kim J, Mathalon DH, et al. (2008) Diffusion
tensor imaging in schizophrenia: Relationship to symptoms. Schizophrenia
Research 98: 157–162.
55. Tang J, Liao Y, Zhou B, Tan C, Liu T, et al. (2010) Abnormal anterior
cingulum integrity in first episode, early-onset schizophrenia: A diffusion tensor
imaging study. Brain Research 1343: 199–205.
56. Mitelman SA, Newmark RE, Torosjan Y, Chu KW, Brickman AM, et al. (2006)
White matter fractional anisotropy and outcome in schizophrenia. Schizophre-
nia Research 87: 138–159.
57. Lipska BK (2004) Using animal models to test a neurodevelopmental hypothesis
of schizophrenia. J Psychiatry Neurosci 29: 282–286.
58. Qiu A, Tuan TA, Woon PS, Abdul-Rahman MF, Graham S, et al. (2010)
Hippocampal-cortical structural connectivity disruptions in schizophrenia: an
integrated perspective from hippocampal shape, cortical thickness, and integrity
of white matter bundles. Neuroimage 52: 1181–1189.
59. Goldman AL, Pezawas L, Mattay VS, Fischl B, Verchinski BA, et al. (2009)
Widespread reductions of cortical thickness in schizophrenia and spectrum
disorders and evidence of heritability. Arch Gen Psychiatry 66: 467–477.
60. Tepest R, Wang L, Miller MI, Falkai P, Csernansky JG (2003) Hippocampal
deformities in the unaffected siblings of schizophrenia subjects. Biol Psychiatry
61. Concha L, Gross DW, Beaulieu C (2005) Diffusion tensor tractography of the
limbic system. AJNR Am J Neuroradiol 26: 2267–2274.
62. Buchsbaum MS, Tang CY, Peled S, Gudbjartsson H, Lu D, et al. (1998) MRI
white matter diffusion anisotropy and PET metabolic rate in schizophrenia.
NeuroReport 9: 425–430.
Tract-Based Analysis on Fornix and Cingulum
PLoS ONE | www.plosone.org11April 2011 | Volume 6 | Issue 4 | e18652