Diffusion Tensor Imaging Reveals White Matter ... · Diffusion Tensor Imaging Reveals White Matter...
Transcript of Diffusion Tensor Imaging Reveals White Matter ... · Diffusion Tensor Imaging Reveals White Matter...
Diffusion Tensor Imaging RevealsWhite Matter Reorganization inEarly Blind Humans
J.S. Shimony1, H. Burton1,2, A.A. Epstein1, D.G. McLaren2,
S.W. Sun1 and A.Z. Snyder1,3
1Mallinckrodt Institute of Radiology, 2Department of Anatomy
and Neurobiology and 3Department of Neurology, Washington
University School of Medicine, St Louis, MO 63110, USA
Multiple functional methods including functional magnetic reso-nance imaging, transcranial magnetic stimulation, and positronemission tomography have shown cortical reorganization in re-sponse to blindness. We investigated microanatomical correlatesof this reorganization using diffusion tensor imaging and diffu-sion tensor tractography (DTT). Five early blind (EB) were com-pared with 7 normally sighted (NS) persons. DTT showed markedgeniculocalcarine tract differences between EB and NS partic-ipants. All EB participants showed evidence of atrophy of thegeniculocortical tracts. Connections between visual cortex and theorbital frontal and temporal cortices were relatively preservedin the EB group. Importantly, no additional tracts were found in anyEB participant. Significant alterations of average diffusivity andrelative anisotropy were found in the white matter (WM) of theoccipital lobe in the EB group. These observations suggest thatblindness leads to a reorganization of cerebral WM and plausiblysupport the hypothesis that visual cortex functionality in blindnessis primarily mediated by corticocortical as opposed to thalamocort-ical connections.
Keywords: blindness, human, magnetic resonance imaging,visual cortex/*physiology
Introduction
Numerous functional imaging studies have demonstrated phys-
iologic responses in visual cortex of blind humans induced
by performance of various tasks that have focused on language
(Buchel and others 1998; Melzer and others 2001; Burton,
Snyder, Conturo, and others 2002; Burton, Snyder, Diamond,
and Raichle 2002; Amedi and others 2003; Burton and others
2003), memory (Amedi and others 2003), mental imagery
(Aleman and others 2001; Vanlierde and others 2003; Lambert
and others 2004), and perceptual processing of tactile
(Sadato and others 1996, 1998, 2002; Gizewski and others
2003; Burton and others 2004, 2005; ) as well as auditory stimuli
(Kujala and others 1995, 2005; Roder and others 1996, 2001;
Liotti and others 1998; Leclerc and others 2000; Weeks and
others 2000; Arno and others 2001; Kujala and others 2005).
These visual cortex responses tend to be strongest and most
extensive in persons who are either congenitally blind or lost
sight soon after birth. Similarly, electrophysiological cross-
modal visual cortex responses have been observed in blind
animals (Rauschecker 1995; Kahn and Krubitzer 2002; Newton
and others 2002). Thus, it is well established that visual cortex
is functionally reorganized in blindness.
The anatomical correlates of visual loss in blind humans
have been relatively unexplored. Abnormalities of the optic
nerves and lateral geniculate nucleus (LGN) have been de-
scribed (Brunquell and others 1984). However, the best
available evidence indicates that the visual cortex is grossly
normal. Breitenseher and others (1998) noted ‘‘abnormal signal’’
in magnetic resonance images (MRI) of the anterior portion of
the optic radiations in 2 of 12 cases. No other study to date
has examined the effects of blindness on the integrity of the
cerebral white matter (WM). The present study examines the
effect of blindness on the cerebral WM using diffusion tensor
imaging (DTI) and diffusion tensor tractography (DTT).
DTI and DTT have emerged during the past several years as
noninvasive techniques to evaluate WM integrity and neuronal
connectivity. DTI (Basser and others 1994) measures the local
diffusion properties of water using a tensor model. The main
quantities of interest are 1) mean apparent diffusion coefficient
(ADC), which measures total molecular motion averaged over
all directions, and 2) anisotropy (Ar), which refers to the degree
to which diffusion exhibits directional (strictly, angular) de-
pendence. Diffusion is characteristically anisotropic in myelin-
ated WM, the axis along which motion is greatest being parallel
to nerve fibers (Chenevert and others 1990; Doran and others
1990; Moseley and others 1990). This anisotropy property
constitutes the basis of DTT, a computational procedure that
reconstructs major fiber bundles in the brain (Jones and others
1998; Conturo and others 1999; Mori and others 1999; Basser and
others 2000; Poupon and others 2000). Before the advent of DTT,
such information could only be obtained by postmortem studies.
DTT is a noninvasive procedure that provides otherwise
unavailable connectivity information. The major limitation is
that DTT is imperfect as a neuroanatomical technique largely
because of reduced ability to track through regions of low signal
to noise and crossing fibers (Virta and others 1999; Pierpaoli
and others 2001). However, this limitation does not preclude
using DTT to reveal population differences in the microscopic
structure of WM, provided that the data are interpreted with
appropriate caution. Here we focus on several WM tracts re-
lated to visual cortex including the geniculocalcarine tract
(GCT). The GCT is among the first structures to be imaged by
DTT (Conturo and others 1999). We contrast DTI and DTT
results in early blind (EB) as compared with normally sighted
(NS) participants. Our results demonstrate that EB humans have
altered diffusion parameters in subcortical WM in the vicinity of
the calcarine sulcus and absent or attenuated geniculocortical
tracts. We interpret these results as supporting the view that
visual cortex function in blind humans is mediated primarily by
corticocortical as opposed to geniculocortical connections.
Methods
SubjectsThe EB group included 5 individuals (2 female) who were born blind
(Table 1). Three of the subjects (EB1, EB2, and EB12) were blind
Cerebral Cortex November 2006;16:1653--1661
doi:10.1093/cercor/bhj102
Advance Access publication December 28, 2005
� The Author 2005. Published by Oxford University Press. All rights reserved.
For permissions, please e-mail: [email protected]
because of retinopathy of prematurity (ROP), a leading cause of
blindness in premature infants. The major risk factor for ROP is high
levels of supplemental oxygen during the neonatal period. Two
individuals (EB4 and EB11) with light sensitivity at the time of test-
ing carried the diagnosis of Leber’s congenital amaurosis (LCA). LCA
is a retinal degenerative disorder of unknown etiology and onset in
infancy. Thus, the cause of blindness in all EB participants was retinal
pathology. None could read print or navigate without aid. We retain
here the EB designation and identification numbers used in our previous
studies (Burton, Snyder, Conturo, and others 2002; Burton, Snyder,
Diamond, and Raichle 2002; Burton and others 2003, 2004, 2005). The
control group included 7 (3 female) NS individuals age matched to
the EB group. All participants provided informed consent following
guidelines approved by the Human Studies Committee of Washington
University and were compensated for their time. Table 1 presents
demographic characteristics of all participants. Except for ophthalmo-
logic causes of blindness, all participants were neurologically normal.
The magnetic resonance structural images showed clinically normal
brain anatomy in all participants.
Image AcquisitionAll imaging was performed on a 1.5-T Siemens Sonata scanner (Erlangen,
Germany). Structural scans included a T1-weighted (T1W) sagittal,
magnetization-prepared rapid gradient echo (MP-RAGE; repetition
time [TR] = 1900 ms, inversion time [TI] = 1100 ms, echo time [TE] =3.93 ms, flip angle = 15�, 1 3 1 3 1.25--mm voxels) and a T2-weighted
(T2W) fast spin echo (TR = 4380 ms, TE = 94 ms, 1 3 1 3 3 mm).
Diffusion-weighted images were acquired in 48 directions, divided into
4 acquisitions of 12 directions each, using a locally modified echo planar
imaging (EPI) sequence (TR = 7000 ms, TE = 113 ms, 2.5-mm isotropic
voxels, 2.5-mm slice gaps, b value = 800 s/mm2). Odd and even slice
scans (73 s each) were interleaved. Thus, 8 scans were needed to
acquire a complete DTI data set. Five complete DTI data sets were
acquired in each participant. The total imaging time was approximately
90 min per participant.
Image RegistrationAll DTT and DTI computations were conducted in untransformed EPI
space thereby avoiding the need to reorient the diffusion data. The
regions of interest (ROI) on which the DTI and DTT results depended
were defined on the MP-RAGE images. Accordingly, the first image-
processing step was to define the spatial relationships between all
images in terms of affine transforms computed by image registration.
Multimodality (e.g., T2W/T1W) image registration was performed
using vector gradient measure (VGM) maximization (Rowland and
others 2005). The first acquired, unsensitized (b = ~0 s/mm2; I0) DTI
volume was registered to the T2W image; stretch and shear were
enabled (12-parameter affine transform) to partially compensate for EPI
distortion. Atlas transformation was computed via the T1W image, which
itself was registered to an atlas representative target produced by
mutual coregistration of MP-RAGE images from 12 normal, young adults.
The atlas target conformed to the Talairach system (Talairach and
Tournoux 1988) as implemented by Lancaster and others (1995).
Algebraic composition of transforms (matrix multiplication) enabled
resampling any data type in register with any other (Ojemann and others
1997). Thus, ROI generated on anatomical images were resampled in
register with the DTI data for purposes of tract selection and DTI
parameter measurement. Figure 1 illustrates the obtained multimodal
image registration accuracy in a representative sighted subject.
Head Motion Correction of the DTI DataEach DTI data set included 52 volumes (48 diffusion sensitized + 4
unsensitized) assembled by collating slices from 2 interleaved scans.
No attempt was made to correct for head motion between odd and
even slice scans. Each 52-volume data set was motion corrected using
a procedure that iteratively cycled through the following steps. 1) Align
each volume to the geometric mean volume of each group of images
sharing the same degree of sensitization (12 3 b = 800 s/mm2, 4 3 I0).
2) Recompute the geometric mean volume. 3) Align each group’s
geometric mean to the first acquired I0 image. 4) Algebraically compose
transforms (volume/group geometric mean with group/I0). Three
cycles through the preceding steps yielded realignments with errors
estimated by internal consistency to be less than 0.1 mm. All transforms
were 9-parameter affine (rigid body + scanner axis stretch) computed
by VGM maximization (Rowland and others 2005). The I0 volumes of
each DTI data set were aligned using conventional intensity correlation
maximization (Snyder 1996). The final, motion-corrected result was
obtained by algebraically composing all transforms (saved from the
iterative procedure) and averaging all data sets after application of the
composed transforms using cubic spline interpolation. The final
resampling step output 52 volumes with doubled in-plane sampling
Table 1Demographic information
ID number Age Sex
Age ofblindnessonset
Lightsensitivitya
Years readingBraille
Cause ofblindnessb
Early 1 54 F 0 � 49 ROPc
Early 2 53 M 0 � 47 ROPEarly 4 39 F 0 þ 31 LCAEarly 11 29 M 0 þ 23 LCAEarly 12 27 M 0 � 22 ROPAverage 40.4SEM 5.7Sighted 1 21 FSighted 2 24 MSighted 3 24 MSighted 4 20 FSighted 5 41 FSighted 6 57 MSighted 7 56 MAverage 34.7SEM 6.2
Note: SEM, standard error of mean.aLight sensitivity was self-reported; EB12 reported having light sensitivity until the age of 13.bCause of blindness was self-reported.cBilateral optic nerve agenesis was determined as the cause of blindness by inspection
of MP-RAGE images.
Figure 1. Demonstration, in a normally sighted individual, of achieved multimodalregistration accuracy. (A) High-resolution T1W (MP-RAGE) structural image. (B) Auto-matic (fuzzy class means based) segmentation of T1W and T2W structural data intocerebrospinal fluid (CSF) (dark gray), GM (light gray), and WM (white). (C) Un-sensitized (averaged I0) component of the diffusion data set. (D) Diffusionanisotropy (Ar). All views show the same parasagittal plane. The red and greenoutlines indicate the outer brain edge and the GM--WM boundary, respectively; thesewere traced (Analyze ROI tool) on the MP-RAGE image and duplicated on the othervolumes. The asterisk indicates a region illustrating diverse contrast mechanisms: Inthe I0 image (C), bright CSF is outside the outer boundary of the brain. Thecorresponding locus in (A) and (B) shows GM bounded by the red and green traces. Inthe anisotropy image (D), both GM and CSF are dark and only WM is bright.
1654 Diffusion Tensor Imaging in Blindness d Shimony and others
density (1.25 3 1.25 3 2.5--mm voxels) in spatial register with the I0volume of the first acquired DTI data set.
Definition of the White Matter ROI Subadjacent tothe Visual CortexConsiderable attention was given to defining an ROI in the WM
subadjacent to primary visual cortex (V1) in both hemispheres of
each participant. The first step was manual segmentation of the V1
cortical gray matter (GM) in the T1W anatomical image (in atlas space)
using Analyze (Mayo Clinic, Rochester, MN). The traced region included
all cortex centered on the calcarine sulcus between the crowns of the
adjacent gyri extending anteroposteriorly from the occipital pole
three-fourth of the way to the parietooccipital sulcus. Care was taken
to avoid extending the region into neighboring occipital sulci. The
region boundary was iteratively refined on multiple views (transverse,
coronal, sagittal). It is likely that the manually segmented cortical
ROI included bordering portions of the secondary visual area (V2) in
addition to V1. Next, the coregistered T1W and T2W structural images
were automatically segmented into regions representing cerebrospinal
fluid, GM, and WM using bispectral fuzzy class means (Bezdek and
others 1993) and manually identified loci in T1W and T2W intensity
space. Artifactual intensity inhomogeneity was corrected prior to
segmentation using a second-order 3-dimensional (3D) polynomial
model of the gain field (Styner and others 2000). The manually defined
V1/V2 ROI and the automatic segmentation results were resampled
in spatial register with the DTI images. Finally, the following automated
steps were taken in sequence: 1) restriction of the ROI to GM voxels,
2) dilation by 2.5 mm in all (x, y, z) directions, and 3) restriction of the
dilated results to WM (Fig. 2).
Definition of the LGN ROIThe LGNs of all NS participants were traced on the T1W structural
images (in atlas space) using Analyze. Visualization of the LGN was less
dependable in the EB participants (see Supplementary Materials).
Accordingly, left and right consensus LGN ROI were created in atlas
space from the LGN tracings of the NS participants (Fig. 3). For each NS
participant, voxels inside the traced LGN were assigned a value of one;
all other voxels were set to zero. These binary-coded images were added
together, and a consensus LGN (for each hemisphere) was created using
a threshold of 2. The same consensus LGN ROI was used in all
participants for the purpose of track selection.
Definition of Additional ROISeveral other WM ROIs were individually selected in anatomical images
in atlas space for the purpose of measuring diffusion parameters (ADC
and Ar). The corpus callosum (CC) extending laterally ±6 mm from the
midline was evenly divided into 4 quadrants along its anterior--posterior
axis. The most posterior quadrant, including the splenium, was evenly
divided into superior and inferior halves (Fig. 5D and E). The inferior
half is known to contain the V1 commissural fibers crossing between
the hemispheres (Dougherty and others 2005). Cubic 216-mm3 ROIs
were selected in the frontal and parietal WM of both hemispheres taking
care to avoid GM.
The following procedure was followed to enable the measurement
of diffusion parameters along the course of the GCT. The GCT could
not be reliably identified in the EB participants (see Results). Therefore,
the regions corresponding to the course of the GCT were determined
from the DTT results in the NS group. Voxels through which GCTs
passed were assigned a value of one in each NS participant; all other
voxels were set to 0. These binary-coded images were transformed to
atlas space. The transformed images then were added together, and
a GCT consensus region was created using a threshold of 3. This
consensus region was divided into 3 equal parts (Fig. 5A, B, and C).
DTI and DTT computationsThe diffusion tensor was calculated using log-linear regression (Basser
and others 1994). Diffusion parameters (ADC and Ar) were evaluated as
detailed in prior publications (Conturo and others 1996; Shimony and
others 1999). The formula for ADC is standard in all laboratories. For
quantitative measures of anisotropy, we used Ar, which is proportional
to relative anisotropy and assumes values in the range 0 to 1.
Tractography was performed using a streamline-type algorithm
(i.e., propagating along the local diffusion tensor principal eigenvector)
very much like that available in widely distributed packages (Xue
and others 1999; Basser and others 2000). The propagation increment
was 0.5 mm. Interpolated tensor field values were evaluated using
tensor basis functions (Aldroubi and Basser 1999; Pajevic and others
2002). All tracks intersecting a regular 1-mm3 grid of seed points
covering the whole brain were computed and stored on disk. Track
termination criteria included Ar < 0.13, radius of curvature (ROC)< 1 mm, and I0 intensity below the parenchymal threshold. The saved
tracks were later selected for display and analysis on the basis of
intersection or termination in selected ROIs (Conturo and others 1999).
All presently reported tracts were selected as intersecting the V1/V2
WM ROI individually obtained in each participant as described earlier.
Quantitative results for the GCT were obtained by counting DTT tracks
intersecting both the individual V1/V2 WM ROI and the consensus LGN
ROI (see above).
Because DTT results are sensitive to small changes in tracking para-
meters, the Ar track termination criterion was systematically explored
Figure 2. Illustration of the V1/V2 ROI obtained in 1 sighted (NS1) and 1 EB (EB7) participant. Sagittal and coronal sections are shown with and without the ROI overlay (red). Thearrows indicate the calcarine sulcus. Note confinement, to within 2.5-mm3 voxel resolution, of these ROIs to subcortical WM. These ROIs were used for tract selection (Figs. 3 and4, Tables 3 and 4) and diffusion parameter measurements (Table 5).
Cerebral Cortex November 2006, V 16 N 11 1655
in the range 0.11--0.15 to verify whether the essential phenomenology
was invariant to this manipulation. Quantitative GCT results obtained by
systematic variation of the Ar and ROC track termination criteria are
reported in the Supplementary Materials.
Results
Anatomical Differences
MP-RAGE structural images revealed absent (EB1) or severely
atrophied (EB7 and EB12) optic nerves/chiasm/tracts in EB
participants. Atrophy in these structures was less extreme in the
2 EB participants who reported light sensitivity (Table 1).
Statistical analysis of the automatically dilated and masked
V1/V2 ROI (Fig. 2) revealed significantly smaller WM but not
GM volumes in EB as compared with NS participants (Table 2).
These differences were not attributable to bias in the manually
outlined GM regions submitted to the automated procedure.
The neurobiological implications of this unanticipated result are
discussed subsequently.
Tractography
The tractography results obtained in all participants were
inspected using 3D Slicer (http://www.slicer.org). The location,
configuration, and thickness of tracts seen in all NS hemispheres
were noted, and norms were determined against which the
EB results were compared. The outcome of this comparison
is reported in the following descriptions and summarized in
Table 3. Features comparable to typical NS results are coded as
‘‘++.’’ The symbol ‘‘+’’ signifies the presence of a tract that was
assessed as noticeably thin in comparison with the range seen in
the NS group. The symbol ‘‘–’’ indicates the complete absence
of a tract.
As viewed in 3D Slicer, the GCT in NS participants emanated
from the V1/V2 ROI as a component of a bundle located lateral
to the occipital horn of the lateral ventricle (Figs. 3 and 4).
Toward the posterior thalamus, the GCT gently curved medially
to enter the region of the LGN. As in our original description
of the GCT (Conturo and others 1999), we did not see a well-
developed loop of Meyer. A typical GCT was observed in 13/14
NS hemispheres. A typical GCT was seen only in 2/10 hemi-
spheres in the EB group (Fig. 3, right hemisphere of EB4 and
left hemisphere of EB11). Corresponding quantitative results,
obtained by counting the number of GCTs intersecting both
the individual V1/V2 WM ROI and the consensus LGN ROI, are
listed in Table 4. The use of 2 well-separated track selection
ROIs effectively eliminates tracks that deviate off course due to
accumulated errors in the locally computed principal diffusivity
orientation. Variation of the track termination criteria changed
the absolute number of GCTs in individuals but did not alter
the EB versus NS proportional results (see Supplementary
Materials). Both EB individuals with any DTT evidence of
a GCT self-reported light sensitivity (Table 1).
Pulvinar or superior colliculus (SC) projections were variably
seen in the NS group; one or both of these features were present
in both hemispheres of all sighted participants (e.g., Fig. 4, NS1,
and Table 3). These bundles, originally lateral to the GCT in
occipital WM, sharply bent medially a few millimeters anterior
to the LGN, crossed the region of the LGN, and projected
toward either the posterior pulvinar or, more ventrally, the SC.
Comparable results were seen in 3/10 EB hemispheres (Table 3)
in the 2 EB participants with self-reported light sensitivity
(Table 1). These DTT results are displayed in Figure 4 (EB4 [only
left side shown] and EB11).
Corticocortical tracks emanating from the V1/V2 ROI were
similarly distributed in all NS participants (e.g., Fig. 4, NS1). One
broad but loosely organized collection of tracks terminated
in the anterior temporal lobe within 2--3 cm of the temporal
pole. A similarly broad collection of tracks terminated in the
orbitofrontal region. A multimillimeter thick, compact bundle
passed through the splenium of the CC to terminate near VI
of the opposite hemisphere. This commissural bundle always
assumed a characteristic horseshoe shape in axial views (Fig. 4).
In contrast to the consistency seen in the NS group, the
corticocortical DTT results in the EB group were variable.
At one extreme, the tractography picture was indistinguishable
from typical NS results (Fig. 4, left hemisphere of EB11). At the
other extreme (EB1), all typical features were bilaterally absent,
Table 2V1/V2 ROI volumetric statisticsa
Left hemisphere Right hemisphere
GM WM GM WM
NS 2013 (269) 3957 (247) 2194 (264) 3664 (227)EB 1809 (233) 2469 (290) 1916 (217) 1982 (169)P valueb — 0.02 — 0.003
aVolumes are in cubic millimeters; mean (standard error of mean).bP value was determined by a two-sided Mann--Whitney test.
Figure 3. GCT tractography results obtained in 2 sighted and 2 EB participants. Trackswere selected as intersecting both individually defined V1/V2 ROI (Fig. 2) and theconsensus LGN region (yellow) established in sighted participants. Selected tracksare shown overlaid on axial slices. Quantitative results for all participants are given inTable 4. NS2 is a typical NS participant, and NS7 is the most abnormal of the NS group.EB4 and EB11 are the 2 blind participants with a detectable GCT.
1656 Diffusion Tensor Imaging in Blindness d Shimony and others
except for projections to the right orbital frontal lobe (Table 3,
not shown in Fig. 4). Generally, the EB DTT outcomes fell
between the two extremes. The EB versus NS differences were
not qualitatively altered by varying the Ar stopping criterion in
the range 0.13 ± 0.02.
Regional Diffusion Tensor Measurements
Table 5 lists regional ADC and Ar measured in selected cerebral
WM ROI. Several WM regions normally related to V1 showed
significant EB versus NS group differences. In all cases, these
differences were in the direction of greater ADC and lower Ar
Figure 4. Tractography results obtained by selection of tracks intersecting individually defined V1/V2 ROI (Fig. 2). Tracks traced to several locations are shown color coded asfollows: LGN (dark blue), pulvinar/SC (light blue), anterior temporal lobe (green), orbitofrontal (yellow), commissural (red). All images show individual DTT results overlaid on theparticipant’s MP-RAGE. Sagittal, axial, and double oblique (inset key) views are shown on successive rows. One NS and 4/5 EB participants are included; EB1 was omitted becauseof a paucity of DTT fibers. Inspection results for all participants are given in Table 3.
Table 3V1/V2 DTT inspection summary
Fiber target NS1 NS2 NS3 NS4 NS5 NS6 NS7 EB1 EB2 EB4 EB11 EB12
Left temporal pole WM þþ þþ þþ þþ þþ þþ þþ � þþ þþ þþ þLeft orbital frontal WM þþ þþ þþ þþ þþ þþ þþ � þþ þþ þþ þþLeft SC/pulvinar þþ þþ þþ þþ þþ þþ þþ � � þþ þþ �Right temporal pole WM þþ þþ þþ þþ þþ þþ þþ � þþ þþ þþ þRight orbital frontal WM þþ þþ þþ þþ þþ þþ þþ þ þþ þþ þþ þþRight SC/pulvinar þþ þþ þþ þþ þþ þþ þþ � � þ � �CC þþ þþ þþ þþ þþ þþ þþ � þþ þþ þþ þþ
Note: �, no tract detected; þ, tract abnormally attenuated; þþ normal tract.
Table 4Geniculocalcarine DTT track countsa
Fiber target NS1 NS2 NS3 NS4 NS5 NS6 NS7 EB1 EB2 EB4 EB11 EB12
Left LGN 1932 409 258 393 438 678 6 0 0 0 164 0Right LGN 656 212 527 1450 38 42 350 0 0 32 0 0
aTrack termination criteria Ar \ 0.13 and ROC\ 1.0 mm.
Cerebral Cortex November 2006, V 16 N 11 1657
in the EB group. Specifically, significant differences were found
in WM juxtaposed to V1/V2 for ADC on the right and for Ar on
the left (Table 5). Significant differences were also found for
ADC and Ar in the most posterior ROI corresponding to the
course of the GCT (Fig. 5C and Table 5). Additionally, signifi-
cantly lower Ar was observed in the ventral half of the splenium
of the CC in the EB group (Fig. 5D and Table 5). No differences
were seen in ROIs not related to VI, that is, frontal/parietal WM
and all other parts of the CC (Fig. 5E and Table 5).
Discussion
DTI has been used to investigate normal and abnormal brain
maturation (Huppi, Maier, and others 1998; Huppi, Warfield,
and others 1998; Neil and others 1998; McKinstry, Mathur,
and others 2002; McKinstry, Miller, and others 2002; Miller and
others 2002; Partridge and others 2004). DTT is a more recent
development that has been used to characterize brain de-
velopment (Berman and others 2005) and normal adult WM
connectivity (Stieltjes and others 2001; Catani and others 2002;
Ciccarelli and others 2003; Jellison and others 2004). The
notion that postnatal experience can affect WM microstructure
is supported by the recent finding that intensive musical
practice leads to measurable DTI changes in deep cerebral
WM (Bengtsson and others 2005). The present work is, to our
knowledge, the first to use either DTI or DTT to investigate the
developmental effects of sensory deprivation.
Limitations of DTT
DTI and DTT both are based on DTI but serve complimentary
scientific purposes. Mean diffusivity and anisotropy are pre-
cisely defined physical properties of tissue. Values obtained
in practice are affected by image noise (Conturo and others
1996), but the measurement procedure is conceptually straight-
forward. ADC and anisotropy conventionally are measured in
targeted ROIs (Pierpaoli and Basser 1996; Shimony and others
1999). In contrast, DTT reconstructs tracks over extended
paths that are not a priori determined. DTT has a less certain
relationship to the underlying anatomy. On the one hand, DTT
frequently generates results that are plausible and apparently
accurate (Stieltjes and others 2001; Catani and others 2002;
Ciccarelli and others 2003; Jellison and others 2004). On the
other hand, DTT is subject to several types of error, including 1)
reduced ability to track through zones of low signal to noise,
low anisotropy (especially below the stopping threshold), and
crossing fibers (Virta and others 1999; Pierpaoli and others
2001), 2) difficulty following tract bifurcations (Basser and
others 2000), and 3) inaccurate determination of principal eigen-
vector orientation (Lori and others 2002; Jones 2003). Thus,
DTT may be reasonably regarded as a technique with a finite
rate of false-negative and false-positive outcomes (Sorensen and
others 2005). In the same vein, the quantitative results reported
in Table 4 should be understood as statistical reflections of
diffusion anisotropy along the course of the GCT, not anatom-
ical fiber counts. We therefore do not assert that our DTT
results provide a complete picture of geniculocortical or V1/V2
cortical connections in either the NS or the EB group. We do,
however, believe that the DTT results, in aggregate, suggest
reduced EB versus NS V1/V2 connectivity with the thalamus.
Summary of Findings
With the preceding DTT caveats in mind, we summarize our
main findings as follows. 1) Blindness leads to altered WM
microanatomy as revealed by DTI and DTT. 2) These abnor-
malities are most apparent in the occipital lobe and ventral
splenium. 3) Tractography suggests that attenuated V1/V2
Table 5DTI directionally invariant regional statistics (see Fig. 5)
ROI ADC (mean [SEM]) P valuea Ar (mean [SEM]) P valuea
NS EB NS EB
Left V1/V2 GM 1.182 (0.055) 1.184 (0.048) — 0.060 (0.003) 0.053 (0.004) —Left V1/V2 WM 0.808 (0.018) 0.854 (0.006) — 0.169 (0.004) 0.141 (0.007) 0.012Left anterior GCT 0.868 (0.033) 0.861 (0.020) — 0.292 (0.012) 0.299 (0.016) —Left middle GCT 0.969 (0.086) 1.061 (0.165) — 0.391 (0.024) 0.283 (0.023) —Left posterior GCT 0.846 (0.043) 0.939 (0.058) — 0.279 (0.012) 0.188 (0.005) 0.003Left frontal WM (�25, 35, 4)b 0.854 (0.019) 0.825 (0.012) — 0.213 (0.016) 0.192 (0.011) —Left parietal WM (�27, �49, 26)b 0.842 (0.020) 0.886 (0.020) — 0.251 (0.02) 0.238 (0.014) —Right V1/V2 GM 1.125 (0.050) 1.171 (0.036) — 0.062 (0.002) 0.057 (0.005) —Right V1/V2 WM 0.791 (0.013) 0.846 (0.006) 0.012 0.185 (0.005) 0.156 (0.013) —Right anterior GCT 0.816 (0.024) 0.818 (0.017) — 0.295 (0.007) 0.305 (0.011) —Right middle GCT 0.878 (0.031) 0.876 (0.058) — 0.378 (0.016) 0.328 (0.017) —Right posterior GCT 0.785 (0.008) 0.857 (0.020) 0.012 0.292 (0.010) 0.199 (0.018) 0.003Right frontal WM (23, 37, 4)a 0.821 (0.021) 0.831 (0.024) — 0.193 (0.016) 0.179 (0.008) —Right parietal WM (27, �47, 26)b 0.815 (0.018) 0.870 (0.026) — 0.235 (0.021) 0.184 (0.011) —Ventral splenium 1.055 (0.035) 1.032 (0.036) — 0.500 (0.017) 0.436 (0.015) 0.048
aP value determined by a two-sided Mann--Whitney test.bTalairach coordinate of ROI center (x, y, z).
Figure 5. Selected ROI used for regional measurement of ADC and Ar (Table 5). A, B,and C: anterior, middle, and posterior thirds of the consensus GCT obtained in the NSgroup. The background slice shows the Talairach atlas representative image at axialplane z = –4. D: inferior half of the splenium. E: other segments of the CC. Thebackground slice is that of a representative NS individual through the midsagittal plane.
1658 Diffusion Tensor Imaging in Blindness d Shimony and others
connectivity predominantly affects thalamocortical connec-
tions. 4) There is no evidence of a DTT feature present in blind
but not in sighted persons. 5) Unanticipated observations
suggest that gross morphological abnormalities may affect the
LGN and occipital lobes of EB individuals (see Supplementary
Materials).
DTT Correlates of Functional Reorganization inBlindness
Reduced thalamocortical connectivity in EB as compared with
sighted people may reflect anatomical loss of fibers or reduced
anisotropy. The present methods cannot distinguish between
these two alternatives. Corticocortical connections between
the occipital, orbitofrontal, and temporal cortices were rela-
tively preserved. These observations constrain explanations
about the probable basis of physiological effects of sensory
deprivation, specifically cross-modal activation in blindness. The
absence of novel thalamocortical connections suggests that
other thalamic nuclei did not convey nonvisual inputs to visual
cortex. Relatively preserved corticocortical connections in the
EB group (Table 3) suggest that functional adaptations in
blindness make use of cross-modal inputs to visual cortex
from other cortical areas. Known corticocortical connections
between lower tier visual cortex and higher level visual areas
and with multisensory parietal and temporal association areas
(Andersen and others 1990; Van Essen and others 1990; Felle-
man and Van Essen 1991; Lewis and Van Essen 2000; Falchier
and others 2002) normally support the flow of information from
lower sensory to higher order and multisensory cortical areas.
Feedback connections exert modulatory effects on lower level
sensory areas (Van Essen and others 1992). These feedback
connections hypothetically convey tactile and auditory input to
visual cortex that, under normal circumstances, may only
modulate the processing of visual information. Sensory depri-
vation may alter the balance between geniculocortical and
corticocortical connections. Experimental support for this idea
is provided by the demonstration of reversible activation in
visual cortex by tactile stimulation after 5 days of visual
deprivation in sighted humans (Pascual-Leone and Hamilton
2001). These findings suggest that competition between visual
and nonvisual inputs is normally present in visual cortex. Such
short-term effects are presumably not due to new anatomical
connections. Thus, in blind individuals, it is plausible that loss of
visual input shifts the competitive synaptic balance toward
processes mediated by input from other cortical areas. We
hypothesize that corticocortical inputs drive visual cortex in
blind people, possibly by enhanced synaptic connections.
This hypothesis, however, applies only to blindness acquired
past a certain developmental stage. Rakic and others demon-
strated retention of basic cytological structure and normal
cortical thickness of area 17 (despite the absence of visual
information) following late gestation binocular and monocular
enucleations in rhesus monkeys (Rakic 1981, 1988; Rakic and
others 1991). In at least 4/5 of the present EB individuals (the
diagnosis in EB1 being somewhat uncertain), the ontogenetic
development of area 17 presumably was normal because the
onset of blindness was perinatal. Normal visual cortex GM
volume in the blind group (Table 2), therefore, is consistent
with the above-mentioned late gestation binocular enucleation
data (Rakic 1988). Extrapolating these results to the present
EB individuals, we would expect that their visual cortex had
a normal complement of cortical cells that supported the
development and maintenance of corticocortical connections
and, hence, the relatively preserved appearance of cortico-
cortical tracts in the EB group (Table 3).
WM Microstructural Changes as Revealed by DTI
The interpretation of the tractography results as suggesting
some abnormality in the EB group is supported by the DTI
measurements. In all regions with significant EB versus NS
diffusion differences, the effect consistently was in the di-
rection of increased diffusivity and reduced anisotropy (Table
5). DTI has limited ability to identify the cellular and molecular
mechanisms underlying the observed effects. However, the
present EB versus NS differences are similar to findings seen in
immaturity (Huppi, Maier, and others 1998; Huppi, Warfield, and
others 1998; Neil and others 1998; Mukherjee and others 2002),
demyelination (Werring and others 1999; Bammer and others
2000; Fillipi and others 2001), and Wallerian degeneration
(Pierpaoli and others 2001).
Gross Anatomical Correlates of Blindness
The reduced voxel counts in subcortical V1/V2 WM (Table 2)
indicate loss of occipital WM volume, presumably reflecting
axonal loss, fiber thinning, or dysmyelination. These gross
morphological changes in blindness deserve further scrutiny.
Anterograde transneuronal degeneration of the LGN is
a commonly reported consequence of enucleations in animals
(Cowan 1970) and humans (Beatty and others 1982). Brunquell
and others (1984) reported that the LGN was gliotic in an
autopsy case of bilateral anophthalmos. They also reported
absent optic nerve/chiasm/tracts in this case. This is consistent
with the degeneration of the LGN in at least 3 of 5 of the EB
participants (those with atrophied or absent peripheral optic
structures) as suggested by the structural images. However, the
extent of LGN atrophy is unclear in the MRI structural data as
gliosis cannot be distinguished from transneuronal degenera-
tion in T1W images.
Summary
DTT results in EB as compared with NS humans suggested that
the main locus of disrupted V1/V2 connectivity involves the
thalamus as opposed to other areas of cortex. Diffusion tensor
measurements (ADC and Ar) showed abnormalities of occipital
WM and the ‘‘visual component’’ of the CC. Additional observa-
tions suggested that EB humans may have degeneration of
the LGN and reduced occipital WM volume. Thus, it appears
that blindness leads to abnormalities of visual cortex-related
WM at both the gross and microstructural levels. At the same
time, the available evidence suggests that the visual cortex itself
is preserved and remains functional, evidently, on the basis of
maintained connections with other areas of the cerebral cortex.
Supplementary Material
Supplementary material can be found at http://www.cercor.
oxfordjournals.org/
Notes
This work was supported by the National Institute of Neurological
Disorders and Stroke NS037237; NS39538; P30NS048056; National
Institutes of Health R01NS047592; National Multiple Sclerosis Society
RG3376; CA1012; and Washington University’s McDonnell Center for
Higher Brain Function.
Cerebral Cortex November 2006, V 16 N 11 1659
Address correspondence to Dr J.S. Shimony, Mallinckrodt Institute
of Radiology, Campus Box 8131, Washington University School of
Medicine, 660 South Euclid Avenue, St Louis, MO 63110, USA.
Email: [email protected].
References
Aldroubi A, Basser PJ. 1999. Reconstruction of vector and tensor fields
from sampled discrete data. Contemp Math 247:1--15.
Aleman A, van Lee L, Mantione MH, Verkoijen IG, de Haan EH. 2001.
Visual imagery without visual experience: evidence from congeni-
tally totally blind people. Neuroreport 12:2601--2604.
Amedi A, Raz N, Pianka P, Malach R, Zohary E. 2003. Early ‘visual’ cortex
activation correlates with superior verbal memory performance in
the blind. Nat Neurosci 6:758--766.
Andersen RA, Asanuma C, Essick G, Siegel RM. 1990. Corticocortical
connections of anatomically and physiologically defined subdivisions
within the inferior parietal lobule. J Comp Neurol 296:65--113.
Arno P, De Volder AG, Vanlierde A, Wanet-Defalque MC, Streel E, Robert
A, Sanabria-Bohorquez S, Veraart C. 2001. Occipital activation by
pattern recognition in the early blind using auditory substitution for
vision. Neuroimage 13:632--645.
BammerR, AugustinM, Strasser-Fuchs S, Seifert T, Kapeller P, Stollberger R,
Ebner FF, Hartung HP, Fazekas F. 2000. Magnetic resonance diffusion
tensor imaging for characterizing diffuse and focal white matter
abnormalities in multiple sclerosis. Magn Reson Med 44:583--591.
Basser PJ, Mattiello J, LeBihan D. 1994. Estimation of the effective self-
diffusion tensor from the NMR spin echo. J Magn Reson B
103:247--254.
Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A. 2000. In vivo fiber
tractography using DT-MRI data. Magn Reson Med 44:625--632.
Beatty RM, Sadun AA, Smith L, Vonsattel JP, Richardson EP Jr. 1982.
Direct demonstration of transsynaptic degeneration in the human
visual system: a comparison of retrograde and anterograde changes.
J Neurol Neurosurg Psychiatry 45:143--146.
Bengtsson SL, Nagy Z, Skare S, Forsman L, Forssberg H, Ullen F. 2005.
Extensive piano practicing has regionally specific effects on white
matter development. Nat Neurosci 8:1148--1150.
Berman JI, Mukherjee P, Partridge SC, Miller SP, Ferriero DM, Barkovich
AJ, Vigneron DB, Henry RG. 2005. Quantitative diffusion tensor MRI
fiber tractography of sensorimotor white matter development in
premature infants. Neuroimage 27:862--871.
Bezdek JC, Hall LO, Clarke LP. 1993. Review of MR image segmentation
techniques using pattern recognition. Med Phys 20:1033--1048.
Breitenseher M, Uhl F, Prayer-Wimberger D, Deecke L, Trattnig S,
Kramer J. 1998. Morphological dissociation between visual pathways
and cortex: MRI of visually deprived patients with congenital
peripheral blindness. Neuroradiology 40:424--427.
Brunquell PJ, Papale JH, Horton JC, Williams RS, Zgrabik MJ, Albert DM,
Hedley-Whyte ET. 1984. Sex-linked hereditary bilateral anophthal-
mos. Arch Opthalmol 102:108--113.
Buchel C, Price C, Frackowiak RS, Friston K. 1998. Different activation
patterns in the visual cortex of late and congenitally blind subjects.
Brain 121(Pt 3):409--419.
Burton H, Diamond JB, McDermott KB. 2003. Dissociating cortical
regions activated by semantic and phonological tasks: a FMRI study
in blind and sighted people. J Neurophysiol 90:1965--1982.
Burton H, McLaren DG, Sinclair RJ. 2005. Reading embossed capital
letters: a fMRI study in blind and sighted individuals. Hum Brain
Mapp 10.1002/hbm.20188.
Burton H, Snyder AZ, Conturo TE, Akbudak E, Ollinger JM, Raichle ME.
2002. Adaptive changes in early and late blind: a fMRI study of Braille
reading. J Neurophysiol 87:589--607.
Burton H, Snyder AZ, Diamond JB, Raichle ME. 2002. Adaptive changes in
early and late blind: a fMRI study of verb generation to heard nouns.
J Neurophysiol 88:3359--3371.
Burton H, Sinclair RJ, McLaren DG. 2004. Cortical activity to vibrotactile
stimulation: an fMRI study in blind and sighted individuals. Hum
Brain Mapp 23:210--228.
Catani M, Howard RJ, Pajevic S, Jones DK. 2002. Virtual in vivo
interactive dissection of white matter fasciculi in the human brain.
Neuroimage 17:77--94.
Chenevert TL, Brunberg JA, Pipe JG. 1990. Anisotropic diffusion in
human white matter: demonstration with MR technique in vivo.
Radiology 177:401--405.
Ciccarelli O, Toosy AT, Parker GJ, Wheeler-Kingshott CA, Barker GJ,
Miller DH, Thompson AJ. 2003. Diffusion tractography based group
mapping of major white-matter pathways in the human brain.
Neuroimage 19:1545--1555.
Conturo TE, Lori NF, Cull TS, Akbudak E, Snyder AZ, Shimony JS,
McKinstry RC, Burton H, Raichle ME. 1999. Tracking neuronal fiber
pathways in the living human brain. Proc Natl Acad Sci USA
96:10422--10427.
Conturo TE, McKinstry RC, Akbudak E, Robinson BH. 1996. Encoding of
anisotropic diffusion with tetrahedral gradients: a general mathe-
matical diffusion formalism and experimental results. Magn Reson
Med 35:399--412.
Cowan WM. 1970. Anterograde and retrograde transneuronal degener-
ation in the central and peripheral nervous system. In: Nauta WJH,
Ebbesson SOE, editors. Contemporary research methods in neuro-
anatomy. Heidelberg: Springer-Verlag. p 217--251.
Doran M, Hajnal JV, Van Brugen N, King MD, Young IR, Bydder GM.
1990. Normal and abnormal white matter tracts shown by MR
imaging using directional diffusion weighted sequences. J Comput
Assisted Tomogr 14:865--873.
Dougherty RF, Ben-Shachar M, Bammer R, Brewer AA, Wandell BA. 2005.
Functional organization of human occipital-callosal fiber tracts. Proc
Natl Acad Sci USA 102:7350--7355.
Falchier A, Clavagnier S, Barone P, Kennedy H. 2002. Anatomical
evidence of multimodal integration in primate striate cortex. J
Neurosci 22:5749--5759.
Felleman DJ, Van Essen DC. 1991. Distributed hierarchical processing in
the primate cerebral cortex. Cereb Cortex 1:1--47.
Fillipi M, Cercignani M, Inglese M, Horsfield MA, Comi G. 2001. Diffusion
tensor magnetic resonance in multiple sclerosis. Neurology
56:304--311.
Gizewski ER, Gasser T, de Greiff A, Boehm A, Forsting M. 2003. Cross-
modal plasticity for sensory and motor activation patterns in blind
subjects. Neuroimage 19:968--975.
Huppi PS, Maier SE, Peled S, Zientara GP, Barnes PD, Jolesz FA, Volpe JJ.
1998. Microstructural development of human newborn cerebral
white matter assessed in vivo by diffusion tensor magnetic resonance
imaging. Pediatr Res 44:584--590.
Huppi PS, Warfield S, Kikinis R, Barnes PD, Zientara GP, Jolesz FA, Tsuji
MK, Volpe JJ. 1998. Quantitative magnetic resonance imaging of
brain development in premature and mature newborns. Ann Neurol
43:224--235.
Jellison BJ, Field AS, Medow J, Lazar M, Salamat MS, Alexander AL. 2004.
Diffusion tensor imaging of cerebral white matter: a pictorial review
of physics, fiber tract anatomy, and tumor imaging patterns. Am J
Neuroradiol 25:356--369.
Jones DK. 2003. Determining and visualizing uncertainty in estimates of
fiber orientation fromdiffusion tensorMRI.MagnResonMed49:7--12.
Jones DK, Simmons A, Williams SCR, Horsfield MA. 1998. Non-invasive
assessment of axonal fiber connectivity in the human brain via
diffusion tensor MRI. Magn Reson Med 42:37--41.
Kahn DM, Krubitzer L. 2002. Massive cross-modal cortical plasticity and
the emergence of a new cortical area in developmentally blind
mammals. Proc Natl Acad Sci USA 99:11429--11434.
Kujala T, HuotilainenM, Sinkkonen J, Ahonen AI, Alho K, HamalainenMS,
Ilmoniemi RJ, Kajola M, Knuutila JE, Lavikainen J, Salonen O, Simola J,
Standertskjold-Nordenstam C, Tiitinen H, Tissari SO, Naatanen R.
1995. Visual cortex activation in blind humans during sound
discrimination. Neurosci Lett 183:143--146.
Kujala T, Palva MJ, Salonen O, Alku P, Huotilainen M, Jarvinen A,
Naatanen R. 2005. The role of blind humans’ visual cortex in auditory
change detection. Neurosci Lett 379:127--131.
Lambert S, Sampaio E, Mauss Y, Scheiber C. 2004. Blindness and brain
plasticity: contribution of mental imagery? An fMRI study. Brain Res
Cogn Brain Res 20:1--11.
Lancaster JL, Glass TG, Lankipalli BR, Downs H, Mayberg H, Fox PT. 1995.
A modality-independent approach to spatial normalization of tomo-
graphic images of the human brain. Hum Brain Mapp 3:209--223.
1660 Diffusion Tensor Imaging in Blindness d Shimony and others
Leclerc C, Saint-Amour D, Lavoie ME, Lassonde M, Lepore F. 2000. Brain
functional reorganization in early blind humans revealed by auditory
event-related potentials. Neuroreport 11:545--550.
Lewis JW, Van Essen DC. 2000. Corticocortical connections of visual,
sensorimotor, andmultimodal processing areas in the parietal lobe of
the macaque monkey. J Comp Neurol 428:112--137.
Liotti M, Ryder K, Woldorff MG. 1998. Auditory attention in the
congenitally blind: where, when and what gets reorganized? Neuro-
report 9:1007--1012.
Lori NF, Akbudak E, Shimony JS, Cull TS, Snyder AZ, Guillory RK,
Conturo TE. 2002. Diffusion tensor fiber tracking of human brain
connectivity: acquisition methods, reliability analysis, and biological
results. NMR Biomed 15:493--515.
McKinstry RC, Mathur A, Miller JH, Ozcan A, Snyder AZ, Schefft GL, Almli
CR, Shiran SI, Conturo TE, Neil JJ. 2002. Radial organization of
developing preterm human cerebral cortex revealed by non-invasive
water diffusion anisotropy MRI. Cereb Cortex 12:1237--1243.
McKinstry RC, Miller JH, Snyder AZ, Mathur A, Schefft GL, Almli CR,
Shimony JS, Shiran SI, Neil JJ. 2002. A prospective, longitudinal
diffusion tensor imaging study of brain injury in newborns. Neurol-
ogy 59:824--833.
Melzer P, Morgan VL, Pickens DR, Price RR, Wall RS, Ebner FF. 2001.
Cortical activation during Braille reading is influenced by early visual
experience in subjects with severe visual disability: a correlational
fMRI study. Hum Brain Mapp 14:186--195.
Miller SP, Vigneron DB, Henry RG, Bohland MA, Ceppi-Cozzio C,
Hoffman C, Newton N, Partridge JC, Ferriero DM, Barkovich AJ.
2002. Serial quantitative diffusion tensor MRI of the premature brain:
development in newborns with and without injury. J Magn Reson
Imaging 16:621--632.
Mori S, Crain BJ, Chacko VP, van Zijl PC. 1999. Three dimensional tracking
of axonal projections in the brain by MRI. Ann Neurol 45:265--269.
Moseley ME, Cohen Y, Kucharczyk J. 1990. Diffusion-weighted MR
imaging of anisotropic water diffusion in cat central nervous system.
Radiology 176:439--445.
Mukherjee P, Miller JH, Shimony JS, Philip JV, Nehra D, Snyder AZ,
Conturo TE, Neil JJ, McKinstry RC. 2002. Diffusion-tensor MR
imaging of gray and white matter development during normal
human brain maturation. Am J Neuroradiol 23:1445--1456.
Neil JJ, Shiran SI, McKinstry RC, Schefft GL, Snyder AZ, Almli CR,
Akbudak E, Aronovitz JA, Miller JP, Lee BC, Conturo TE. 1998. Normal
brain in human newborns: apparent diffusion coefficient and
diffusion anisotropy measured by using diffusion tensor MR imaging.
Radiology 209:57--66.
Newton JR, Sikes RW, Skavenski AA. 2002. Cross-modal plasticity after
monocular enucleation of the adult rabbit. Exp Brain Res 144:
423--429.
Ojemann JG, Akbudak E, Snyder AZ, McKinstry RC, Raichle ME, Conturo
TE. 1997. Anatomic localization and quantitative analysis of gradient
refocused echo-planar fMRI susceptibility artifacts. Neuroimage
6:156--167.
Pajevic S, Aldroubi A, Basser PJ. 2002. A continuous tensor field
approximation of discrete DT-MRI data for extracting microstruc-
tural and architectural features of tissue. J Magn Reson 154:85--100.
Partridge SC, Mukherjee P, Henry RG, Miller SP, Berman JI, Jin H, Lu Y,
Glenn OA, Ferriero DM, Barkovich AJ, Vigneron DB. 2004. Diffusion
tensor imaging: serial quantitation of white matter tract maturity in
premature newborns. Neuroimage 22:1302--1314.
Pascual-Leone A, Hamilton R. 2001. The metamodal organization of the
brain. Prog Brain Res 134:427--445.
Pierpaoli C, Barnett A, Pajevic S, Chen R, Penix L, Virta A, Basser PJ. 2001.
Water diffusion changes in Wallerian degeneration and their de-
pendence on white matter architecture. Neuroimage 13:1174--1185.
Pierpaoli C, Basser PJ. 1996. Toward a quantitative assessment of
diffusion anisotropy. Magn Reson Med 36:893--906.
Poupon C, Clark CA, Frouin V, Regis J, Bloch I, LeBihan D, Mangin J.
2000. Regularization of diffusion-based direction maps for the
tracking of brain white matter fasciculi. Neuroimage 12:184--195.
Rakic P. 1981. Development of visual centers in the primate brain depends
on binocular competition before birth. Science 214:928--931.
Rakic P. 1988. Specification of cerebral cortical areas. Science
241:170--176.
Rakic P, Suner I, Williams RW. 1991. A novel cytoarchitectonic area
induced experimentally within the primate visual cortex. Proc Natl
Acad Sci USA 88:2083--2087.
Rauschecker JP. 1995. Compensatory plasticity and sensory substitution
in the cerebral cortex. Trends Neurosci 18:36--43.
Roder B, Rosler F, Hennighausen E, Nacker F. 1996. Event-related
potentials during auditory and somatosensory discrimination in
sighted and blind human subjects. Brain Res Cogn Brain Res 4:77--93.
Roder B, Rosler F, Neville HJ. 2001. Auditory memory in congenitally
blind adults: a behavioral-electrophysiological investigation. Brain
Res Cogn Brain Res 11:289--303.
Rowland DJ, Garbow JR, Laforest R, Snyder AZ. 2005. Registration
of [18F]FDG microPET and small-animal MRI. Nucl Med Biol
32:567--572.
Sadato N, Okada T, Honda M, Yonekura Y. 2002. Critical period for
cross-modal plasticity in blind humans: a functional MRI study.
Neuroimage 16:389--400.
Sadato N, Pascual-Leone A, Grafman J, Deiber MP, Ibanez V, Hallett M.
1998. Neural networks for Braille reading by the blind. Brain
121(Pt 7):1213--1229.
Sadato N, Pascual-Leone A, Grafman J, Ibanez V, Deiber MP, Dold G,
Hallett M. 1996. Activation of the primary visual cortex by Braille
reading in blind subjects. Nature 380:526--528.
Shimony JS, McKinstry RC, Akbudak E, Aronovitz JA, Snyder AZ, Lori NF,
Cull TS, Conturo TE. 1999. Quantitative diffusion-tensor anisotropy
brain MR imaging: normative human data and anatomic analysis.
Radiology 212:770--784.
Snyder AZ. 1996. Difference image vs. ratio image error function forms
in PET-PET realignment. In: Bailey D, Jones T, editors. Quantifica-
tion of brain function using PET. San Diego, CA: Academic Press.
p 131--137.
Sorensen AG, Wang R, Benner T, Makris N. 2005. An approach to
validation of diffusion MRI-based white matter tractography. In:
Proceedings of the Thirteenth Annual Meeting of the International
Society for Magnetic Resonance in Medicine; 2005 7-13 May, Miami
Beach, FL. p 224.
Stieltjes B, Kauffman WE, van Zijl PC, Fredericksen K, Pearlson GD,
Solaiyappan M, Mori S. 2001. Diffusion tensor imaging and axonal
tracking in the human brainstem. Neuroimage 14:732--735.
Styner M, Brechbuhler C, Szekely G, Gerig G. 2000. Parametric estimate
of intensity inhomogeneities applied to MRI. IEEE Trans Med
Imaging 19:153--165.
Talairach J, Tournoux P. 1988. Coplanar stereotaxic atlas of the human
brain. New York: Thieme Medical.
Van Essen DC, Anderson CH, Felleman DJ. 1992. Information processing
in the primate visual system: an integrated systems perspective.
Science 255:419--423.
Van Essen DC, Felleman DJ, DoYoe EA, Olavaria J, Knierim J. 1990.
Modular and hierarchical organization of extrastriate visual cortex
in the macaque monkey. Cold Spring Harbor Symp Quant Biol
55:679--696.
Vanlierde A, De Volder AG, Wanet-Defalque MC, Veraart C. 2003.
Occipito-parietal cortex activation during visuo-spatial imagery in
early blind humans. Neuroimage 19:698--709.
Virta A, Barnett A, Pierpaoli C. 1999. Visualizing and characterizing
white matter fiber structure and architecture in the human
pyramidal tract using diffusion tensor MRI. Magn Reson Imaging
17:1121--1133.
Weeks R, Horwitz B, Aziz-Sultan A, Tian B, Wessinger CM, Cohen LG,
Hallett M, Rauschecker JP. 2000. A positron emission tomographic
study of auditory localization in the congenitally blind. J Neurosci
20:2664--2672.
Werring DJ, Clark CA, Barker GJ, Thompson AJ, Miller DH. 1999.
Diffusion tensor imaging of lesions and normal appearing white
matter in multiple sclerosis. Neurology 52:1626--1632.
Xue R, van Zijl PC, Crain BJ, Solaiyappan M, Mori S. 1999. In vivo three-
dimensional reconstruction of rat brain axonal projections by
diffusion tensor imaging. Magn Reson Med 42:1123--1127.
Cerebral Cortex November 2006, V 16 N 11 1661