Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
named after the English botanist Robert Brown, who in 1827observed the constant movement of minute particles sus-pended within grains of pollen.* We now know that molecularmotion is affected by the properties of the medium in which itoccurs and that diffusion within biological tissues reflectsboth tissue structure and architecture at the microscopiclevel. Equal, or isotropic, diffusion occurs when a mediumdoes not restrict molecular motion, as would be the case withcerebrospinal fluid. Skewed, or anisotropic, diffusion, seen incrystals and polymer films, is not equal in all directions. DTI
REVIEW
Diffusion Tensor Imaging and Its Applicationto Neuropsychiatric Disorders
Marek Kubicki, MD, PhD, Carl-Fredrik Westin, PhD, Stephan E. Maier, MD, PhD, Hatsuho Mamata, MD, PhD,Melissa Frumin, MD, Hal Ersner-Hershfield, BA, Ron Kikinis, MD, Ferenc A. Jolesz, MD, Robert McCarley, MD,
and Martha E. Shenton, PhD
Magnetic resonance diffusion tensor imaging (DTI) is a new technique that can be used tovisualize and measure the diffusion of water in brain tissue; it is particularly useful for evalu-ating white matter abnormalities. In this paper, we review research studies that have ap-plied DTI for the purpose of understanding neuropsychiatric disorders. We begin with a dis-cussion of the principles involved in DTI, followed by a historical overview of magneticresonance diffusion-weighted imaging and DTI and a brief description of several differentmethods of image acquisition and quantitative analysis. We then review the applicationof this technique to clinical populations. We include all studies published in English fromJanuary 1996 through March 2002 on this topic, located by searching PubMed and Medlineon the key words “diffusion tensor imaging” and “MRI.” Finally, we consider potential fu-ture uses of DTI, including fiber tracking and surgical planning and follow-up. (HARVARD REV
PSYCHIATRY 2002;10:324–336.)
From the Clinical Neuroscience Division, Laboratory of Neuroscience,Boston VA Health Care System—Brockton Division, Brockton, Mass.(Drs. Kubicki, Frumin, McCarley, and Shenton; Mr. Ersner-Hershfield); the Department of Radiology, Brigham and Women’sHospital, Boston, Mass. (Drs. Kubicki, Westin, Maier, Mamata, Kiki-nis, Jolesz, and Shenton); and the Departments of Psychiatry (Drs.Kubicki, Frumin, McCarley, and Shenton) and Radiology (Drs. Ku-bicki, Westin, Maier, Mamata, Kikinis, Jolesz, and Shenton), Har-vard Medical School, Boston, Mass.
Original manuscript received 14 February 2002; revised manuscriptreceived 29 May 2002, accepted for publication 6 June 2002.
This work was supported, in part, by grants from the National Al-liance for Research on Schizophrenia and Depression (Drs. Kubickiand Frumin), the Grable Foundation (Dr. Kubicki), the National In-stitutes of Health (K02 MH 01110 and R01 MH 50747 to Dr. Shenton,R01 NS 39335 to Dr. Maier, R01 MH 40799 to Dr. McCarley), andthe National Center for Research Resources (R01 RR 11747 to Dr. Kiki-nis, P41 RR 13218 to Drs. Jolesz and Westin); Department of Veter-ans Affairs Merit Awards (Drs. Shenton and McCarley); and a VAPsychiatry/Neuroscience Research Fellowship Award (Dr. Frumin).
324
Diffusion tensor imaging (DTI) is an exciting recent techniquein neuroimaging that affords a unique opportunity to quan-tify the diffusion of water in brain tissue. It is based upon thephenomenon of water diffusion known as Brownian motion,
Reprint requests: Martha E. Shenton, PhD, Department of Psychiatry–116A, VA Boston Healthcare System—Brockton Division, 940 BelmontSt., Brockton, MA 02301 (e-mail: [email protected]).
© 2002 President and Fellows of Harvard College
*It was long thought that Brown observed the movement of pollengrains suspended in water. Many now believe that he observed themovement of particles suspended within the grains of pollen. See, forexample, BJ Ford, Brownian movement in Clarkia pollen: a repriseof the first observations. The Microscope 1992;40:235–41 (avail-able on the World Wide Web at: http://www.sciences.demon.co.uk/wbbrowna.htm).
measures diffusion properties and consequently allows spa-tial description of the medium under study.
Taking advantage of the fact that diffusion is not uniformthroughout the brain (differing, for example, between graymatter, white matter, and cerebrospinal fluid), researcherscan employ DTI to evaluate tissue characteristics. The tech-nique is particularly useful in the study of white matter tractsin the brain since the mobility of water is restricted perpen-dicular to the axons oriented along the fiber tracts (aniso-tropic diffusion). This is due to the concentric structure ofmultiple tightly packed myelin membranes wrapped aroundthe axon fibers. Although myelination is not essential for dif-fusion anisotropy of nerves (see studies on nonmyelinatedgarfish olfactory nerves1 and on neonate brains prior to theappearance of myelin2,3), myelin is generally assumed to bethe major barrier to diffusion in white matter tracts.
DTI evolved from earlier studies using diffusion-weightedimaging (DWI), a magnetic resonance imaging (MRI) tech-nique in which a single field gradient pulse is applied duringimage acquisition, allowing quantitative measurement of wa-ter diffusion.4 Displacement of water molecules (diffusion)causes randomization of the nuclear magnetic resonance spinphase, which, in turn, results in signal reduction. The amountof reduction provides a quantitative measure of the diffusionin the gradient direction; thus, only diffusion in the directionof this particular gradient can be detected. Since diffusion isa three-dimensional process, three orthogonal measures areneeded to calculate the mean diffusivity for each voxel.
DTI was developed for true multidimensional assessmentof diffusion data in vivo.5,6 In contrast to DWI, DTI measuresat least six different gradient directions. The diffusion datafor each voxel, represented as a 3 × 3 matrix, comprise a dif-fusion tensor (see Figure 1). In isotropic media, where diffu-sion along the three main axes is equal, the diffusion tensor issymmetrical in all directions and is visualized as a sphere. Inanisotropic media, where the diffusion is different along eachaxis, the diffusion tensor is visualized as an ellipsoid, withits longest axis indicating the greatest of the so-called prin-cipal directions of diffusion. The shape of the tensor ellipsoiddepends on the strength of the diffusion along the three prin-cipal directions (i.e., its eigenvectors). Within myelinatedwhite matter fiber tracts, the greatest principal direction ofdiffusion will always indicate the axonal trajectory, since per-pendicular diffusion is restricted by myelin sheathing. Theshape of the tensor ellipsoid therefore provides qualitativeand quantitative measures of white matter tracts withinthe brain.
DWI AND ITS ACQUISITION IN THE BRAIN
DWI was introduced in 1986 by Le Bihan and colleagues.4
From the beginning, however, the widespread application ofthis technique to clinical studies was greatly impeded by tech-
nical constraints, the most important being motion sensitiv-ity, which can cause severe ghosting artifacts or complete sig-nal loss. In attempts to observe molecular displacement inmicrometers, it is no surprise that motion of any sort, evenunavoidable involuntary head movements or pulsations ofblood in the brain tissue, interfere with measurement. Theproblem is even more serious when scans must be obtainedfrom, for example, a disoriented and confused stroke victim,who may move his or her head excessively. These limitationswere a major incentive for the development of faster se-quences that are more robust in the face of bulk motion.
The development of diffusion-sensitive pulse sequencesfollowed two basic directions: echo-planar imaging methods,7
which capture a complete image within a single shot, andnavigator methods,8 which acquire images in multiple shots,with each shot employing “navigator MR signals” to detectand correct the bulk motion. Although single-shot methodsare extremely robust, the elevated sensitivity to magneticfield inhomogeneities inherent in these techniques may leadto image-distortion artifacts, such as susceptibility artifacts,occurring in areas exhibiting large variations in magneticsusceptibility (e.g., interfaces between air, bone, and braintissue), and chemical shift artifacts, caused by the differencein chemical properties of fat and water. Moreover, spatial res-olution is limited, and signal averaging may be necessary.Navigator methods, on the other hand, permit excellent spa-tial resolution with a minimum of image-distortion artifactsand high signal-to-noise ratio, but they are not as robust andrequire acquisition times of 10 minutes or more. Further-more, cardiac gating—that is, synchronization of slice acqui-sition with heart rate—must be used, which makes the tech-nique less attractive in a routine clinical setting. Recently,
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 325
FIGURE 1. Difference between unrestricted (isotropic) and restric-ted (anisotropic) diffusion within the brain. The shape of the tensorellipsoid is determined by the strength of the diffusion along threeprincipal directions (its eigenvectors). In nonrestrictive media suchas cerebrospinal fluid, where diffusion is equal in all directions, thetensor can be visualized as a sphere. In restrictive media such aswhite matter, where diffusion is different in all directions, the tensorcan be visualized as an ellipsoid.
researchers have proposed several new techniques (diffusion-weighted radial acquisition of data,9 line-scan diffusion im-aging,10 slab-scan diffusion imaging13) that avoid susceptibil-ity and chemical shift artifacts and allow for resolution higherthan that obtained with echo-planar imaging.
QUANTITATIVE REPRESENTATION OF DIFFUSION IN DWI
As noted previously, the measurement of water diffusion intissues is based on probing the movement of water moleculeswithin the tissue environment. In pure liquids, such as wa-ter, individual molecules are in constant motion in everydirection due to random (Brownian) motion. In tissues, how-ever, various tissue components (larger molecules, intra-cellular organs, membranes, cell walls, and so on) restrict theBrownian motion. In cerebrospinal fluid and many tissues(liver and cerebral gray matter, for example), when averagedover the macroscopic scale of image voxels, this restriction isidentical in every direction—the diffusion is isotropic. In somevery structured tissues, however, such as muscle or cerebralwhite matter, cellular arrangement shows a preferred direc-tion of water diffusion that is largely uniform across the en-tire voxel—the diffusion is anisotropic. The diffusion coeffi-cient is a measure of this molecular motion, and it can bedetermined by applying consecutive magnetic field gradientpulses and then measuring the change between the imagesacquired. Each gradient is typically applied for several tens ofmilliseconds, during which time the average water moleculein brain tissues may migrate 10 or more micrometers in arandom direction. The irregularity of the motion entails a sig-nal loss that can be used to quantify the diffusion constant.This MR measurement, however, fails to differentiate diffu-sion-related motion from blood flow, perfusion, bulk tissue,or tissue pulsation motions. Thus, the diffusion value ob-tained with this technique is not an actual diffusion coeffi-cient, but only an apparent diffusion coefficient (ADC).
DIFFUSION TENSOR IMAGING
The concept of a diffusion tensor was introduced to the field ofMR diffusion imaging by Basser and colleagues in 1994.6 Itis a construct adapted from physics and engineering, where itis employed to describe tension forces in solid bodies with anarray of three-dimensional vectors.
The particular tensors used to describe diffusion can befurther conceptualized and visualized as ellipsoids. The threemain axes of the ellipsoid describe an orthogonal coordinatesystem. The directions of the main axes represent the so-called eigenvectors; their length, the so-called eigenvalues ofthe tensor. In DTI, a tensor that describes diffusion in all spa-tial directions is calculated for each voxel. The longest mainaxis of the diffusion ellipsoid represents the value and the di-rection of maximum diffusion, whereas the shortest axis rep-
resents the value and direction of minimum diffusion. If thethree eigenvalues are equal, then the diffusion is said to beisotropic, and the diffusion tensor can be visualized as asphere. If they are unequal, then the diffusion is said to beanisotropic, and the diffusion tensor can be visualized moreas an ellipsoid, as would be the case for myelin sheaths (seeFigure 1). To estimate the diffusion tensor, at least six mea-surements (taken from different gradient directions) areneeded, in addition to the baseline image data.
White matter fiber tracts consist of a large number ofdensely packed myelinated axons. Because the movement ofwater molecules within this myelinated white matter is sub-stantially restricted perpendicular to the longitudinal axesof the axons, the longest main axis of the diffusion ellipsoidis much larger than the other two and coincides with thedirection of the fibers. Following Westin and colleagues’ geo-metric classification of the diffusion tensor using linear,planar, and spherical measures,12 this type of anisotropicallyrestricted diffusion is termed “linear diffusion.” “Planar dif-fusion” refers to diffusion restricted in one direction only andunrestricted in the other two—for example, between layersof tissue.
The above-mentioned basic ellipsoid model is idealized anddoes not necessarily reflect the true diffusion behavior en-countered in real tissues. For example, at nerve-fiber-tractcrossings, the ellipsoid tensor model fails, since each fibertract registers a principal direction of diffusion. Acquisitionprotocols that measure diffusion in a large number of direc-tions allow for a better description of the complex directionaldiffusion behavior at fiber-tract crossings and in other het-erogeneously organized tissue structures.
Data from DTI can be analyzed in several ways. The mostgeneral approach is to characterize the overall displacement ofthe molecules (average ellipsoid size) by determining mean dif-fusivity. To do so, the trace of the diffusion tensor,4 which is cal-culated as the sum of the eigenvalues of the tensor, is employed.This sum is divided by three to calculate mean diffusivity.
Several measures have been introduced to describeanisotropic diffusion. To be useful, such measures must be in-dependent of the orientation of the diffusion ellipsoid andthus provide information relevant to the specific tissue type.The most commonly used measures, proposed by Basser andPierpaoli,13 are relative anisotropy (RA), a normalized stan-dard deviation representing the ratio of the anisotropic partof the tensor to its isotropic part; fractional anisotropy (FA), ameasure of the fraction of the magnitude of the tensor thatcan be ascribed to the anisotropic diffusion; and volume ra-tio, a measure representing the ratio of the ellipsoid volumeto the volume as a sphere of radius l. These and otheranisotropy indices are summarized in Table 1. Such indicesmeasure the diffusion within each voxel (intravoxel diffusion)separately. A second type of anisotropy measure has been in-troduced to describe the intervoxel coherence of the tensors in
326 Kubicki et al.Harvard Rev Psychiatry
November/December 2002
the neighboring voxels. The latter measures, summarized inTable 2, better reflect fiber organization and orientation atthe macroscopic level.
Like quantifying diffusion tensor anisotropy, displayingtensors in three dimensions also poses a problem. Severalmethods have been proposed for visualizing the three-dimensional information contained in DTI data. These in-clude using the octahedra in each pixel;14 color maps,15 wheredifferent intensities of the three colors indicate the size andthe ADC in each of the three Cartesian directions;16 and bluelines to represent the in-plane component of the principal dif-fusion direction, along with a color-coded out-of-plane com-ponent17 (see Figure 2).
Clinical Applications in Neuropsychiatric DisordersThe phenomenon of restricted diffusion is of particular inter-est to studies that evaluate the integrity of white matter fibertracts, as noted above. Based on geometry and the degree ofanisotropy loss, white matter tract pathology, such as dislo-cation, swelling, infiltration, and disruption, can be docu-mented. In addition, the cross-sectional sizes of these path-ways yield a quantitative measure of connectivity betweendifferent brain regions. For example, disruptions in connec-tivity—and, in some cases, subsequent reorganizationof nerve pathways—resulting from physical trauma or is-chemia, brain tumor, multiple sclerosis (MS), infection withhuman immunodeficiency virus (HIV), schizophrenia, or de-generative or metabolic diseases might be visualized and
quantified. A loss of connectivity between brain regions asmeasured by DTI could indicate developmental pathology, ax-onal damage, demyelination, and/or disruption of fiber tracts.
Brain ischemia. Evaluation of ischemia is one of the earli-est, most important, and most widely used clinical applica-tions of DWI. So far, DWI is the most sensitive in vivo methodfor detecting acute ischemia.18,19 It also allows for the dis-tinction between old and new strokes19,20 and helps to differ-entiate early stroke from other focal brain processes mimick-ing stroke on conventional MRI.21,22 DWI studies23–25 showthat diffusion parameters decrease in the acute stage,“pseudonormalize” in the subacute phase, and increase in thechronic stage of the stroke. Unfortunately, because DWI doesnot reflect the spatial organization of the fiber tracts, it fails todetect long-term white matter changes, either in the closevicinity of the lesion or remote from it.
DTI, on the other hand, is more sensitive to the organi-zation and orientation of the fiber tracts, and research26 al-ready shows better correlation between clinical status andanisotropy indices (e.g., diffusion anisotropy remains de-creased in the subacute stage, even after diffusion coefficientnormalization [between 1 and 3 weeks.]). DTI data are moresensitive to changes in fiber-tract organization (i.e., axonalloss, degeneration, or incomplete remyelination) followinga stroke. Moreover, changes in anisotropy can be detectedseveral months after the stroke—and within the fiber tractsremote from the stroke (e.g., in the corticospinal tract)—
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 327
TABLE 1. Anisotropy Indices Used in Clinical Studies
Index Study Description
Anisotropy minor, anisotropy major Shimony et al.77 Measures describing the variation in ellipsoid shape along the majorand minor axes
Eccentricity Tievsky et al.41 Measure describing the ellipsoid nature of the principal and minoraxes with reference to a sphere
Fractional anisotropy Basser & Pierpaoli13 Measure of the fraction of the magnitude of the tensor that can be ascribed to anisotropic diffusion
Geometric measures of anisotropy Westin et al.78 Measures classifying the ellipsoid’s closeness to a line, a plane, or asphere (linear, planar, or spherical diffusion)
Gamma variate anisotropy Armitage & Bastin79 A measure formulated in relation to noise (less sensitive to noise thanother anisotropy indices)
Relative anisotropy Basser & Pierpaoli13 Normalized standard deviation representing the ratio of theanisotropic portion of the tensor to its isotropic portion
Total anisotropy Shimony et al.77 Coefficient of variation determined on the basis of the second moment(variance) of diffusion; similar to relative anisotropy except for ascaling factor of 21⁄2, which places total anisotropy on the absoluteanisotropy scale
Ultimate anisotropy Ulug & Van Zihl76 Ratio of the diffusion ellipsoid’s volume to its surface area; does notrequire tensor diagonalization (as opposed to relative and fractionalanisotropy, which do)
Volume ratio Basser & Pierpaoli13 Measure representing the ratio of the ellipsoid volume to the volume asa sphere with a radius of 1
presumably marking fiber-tract degeneration (walleriandegeneration).27,28
A recent DTI study29 has also demonstrated that ischemicstroke damage in white matter occurs earlier and is more se-vere than previously inferred from DWI investigations. Table3 summarizes all clinical DTI studies published throughMarch 2002 on stroke and other neuropsychiatric conditions.
Traumatic brain injury. As with stroke, DWI has playeda major role in the early detection of brain changes following
traumatic injury. The early increase in ADC seen on DWI isusually attributable to vasogenic edema, whereas later de-creases in ADC (despite ongoing increase of intracranial pres-sure) are usually attributable to cytotoxic edema,30 believed toplay a major role in posttraumatic brain swelling.31
DTI allows a closer investigation of the specific fiber tractsaffected, as well as the ability to monitor the degenerationprocess following injury. Specifically, DTI performed severalmonths after an internal capsule focal brain injury32 demon-strated full recovery and preservation of the structural in-tegrity and orientation in the posterior capsular limb and dis-rupted structure in the anterior limb on the injured side,which correlated with the functional motor deficits revealedby the functional MRI. Moreover, in another DTI study33 apatient who had right frontotemporal brain injury and im-paired memory revealed increased diffusion traces in rightfrontal, temporal, and occipital lobes as well as diffusion
328 Kubicki et al.Harvard Rev Psychiatry
November/December 2002
TABLE 2. Intervoxel Diffusion Indices Describing Local Coherence between Tensors
Index Study Description
Correlation measure of Basser & Pierpaoli13 A measure assessed by applying the convolution averaging procedure, organization weighting the scalar matrix products in neighboring voxels more heavily than
those in distant onesGeometric measures of Westin et al.12 Measures obtained by locally averaging the tensors with a low-pass filter and
weighted average tensor applying Westin’s geometric measuresIntervoxel coherence Pfefferbaum et al.50 A measure obtained by averaging the angle between the eigenvector of the largest
eigenvalue of a given voxel and those of its eight neighborsLattice index of anisotropy Pierpaoli & Basser16 A measure that averages diffusion tensors in neighboring voxels, decreasing the
bias and variance of estimated diffusion anisotropy
FIGURE 3. Three-dimensional tractography of a normal subject,showing the anterior (white) and posterior (blue) portions of the cor-pus callosum as well as the left and right (yellow and green) corti-cospinal tracts. These tracts pass through an axial section of the lat-eral ventricles.FIGURE 2. Visualization of diffusion tensors. The blue lines repre-
sent the in-plane component of the principal diffusion direction; theother colors show the magnitude of the out-of-plane component, withorange/red indicating maximal diffusion. The white and green ar-rows point to the corpus callosum and the anterior commissure, re-spectively, two major fiber tracts with the largest in-plane diffusioncomponent, while the blue and pink arrows indicate the cingulateand uncinate fasciculi, fibers with the biggest out-of-plane diffusioncomponent.
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 329
TAB
LE
3.C
linic
al A
pp
licat
ion
s o
f Dif
fusi
on
Ten
sor
Imag
ing
n (
pat
ien
ts/
Con
dit
ion
Stu
dy
con
trol
s)M
easu
re(s
)F
ind
ings
*
Str
oke
Jon
es e
t al
.809/
10Tr
, FA
Dec
reas
e in
an
isot
ropy
, in
crea
se in
dif
fusi
vity
wit
hin
th
e la
cun
ar in
farc
tion
Sor
enso
n e
t al
.8150
/0FA
, Tr,
eig
enva
lues
Red
uct
ion
of a
nis
otro
py w
ith
in w
hit
e m
atte
r in
acu
te c
ereb
ral s
trok
e, a
ttri
buta
ble
to
of t
he
ten
sor
chan
ges
in t
he
firs
t an
d se
con
d ei
gen
valu
es (a
lign
ed w
ith
th
e lo
ng
axes
of t
he
wh
ite
mat
ter
fibe
r tr
acts
)Z
elay
a et
al.26
6/0
Tr, F
A, L
IM
onot
onic
an
d si
gnifi
can
t de
crea
se in
an
isot
ropy
mea
sure
s w
ith
in t
he
isch
emic
lesi
on
from
th
e ac
ute
to
the
chro
nic
sta
geM
ukh
erje
e et
al.29
12/0
Isot
ropi
c di
ffu
sion
Mor
e-se
vere
red
uct
ion
in d
iffu
sion
mea
sure
s w
ith
in w
hit
e m
atte
r th
an w
ith
in g
ray
coef
fici
ent
(sim
ilar
mat
ter
in a
cute
to
earl
y su
bacu
te s
trok
eto
Tr)
, dif
fusi
onan
isot
ropy
Wer
rin
g et
al.27
5/0
Tr, F
AR
edu
ced
anis
otro
py 6
mo
afte
r ce
rebr
al in
farc
tion
wit
hin
th
e co
rtic
ospi
nal
tra
ct, r
emot
e fr
om t
he
lesi
onP
ierp
aoli
et
al.28
7/0
Tr, F
A, L
IR
edu
ced
anis
otro
py w
ith
in t
he
enti
re fi
ber
trac
t af
fect
ed b
y th
e la
cun
ar in
farc
tion
, du
e to
wal
leri
an d
egen
erat
ion
Bra
in in
jury
Wer
rin
g et
al.32
1/5
FAD
ecre
ased
an
isot
ropy
wit
hin
th
e an
teri
or li
mb
of t
he
inte
rnal
cap
sule
, wh
ich
cor
rela
ted
wit
h t
he
fun
ctio
nal
mot
or d
efici
ts r
evea
led
by fM
RI
seve
ral m
onth
s af
ter
the
inte
rnal
ca
psu
le fo
cal b
rain
inju
ryW
iesh
man
n e
t al
.331/
0FA
, Tr
Incr
ease
d di
ffu
sivi
ty in
th
e ri
ght
fron
tal,
tem
pora
l, an
d oc
cipi
tal l
obes
, as
wel
l as
re-
duce
d an
isot
ropy
in t
he
righ
t op
tic
radi
atio
n a
nd
forc
eps
occi
pita
lis,
in a
pat
ien
t w
ith
ri
ght
fron
tote
mpo
ral b
rain
inju
ry s
uff
erin
g fr
om im
pair
ed m
emor
yJo
nes
et
al.82
5/0
TrR
edu
ced
diff
usi
vity
in t
he
peri
pher
y of
a c
ereb
ral c
ontu
sion
des
pite
a n
egat
ive
MR
IB
rain
tu
mor
Bas
tin
et
al.83
6/0
TrA
fter
dex
amet
has
one
trea
tmen
t, d
ecre
ase
in t
race
wit
hin
tu
mor
an
d su
rrou
ndi
ng
edem
atou
s ti
ssu
eW
iesh
man
n e
t al
.341/
20FA
, Tr
Dev
iati
on o
f fibe
rs in
nor
mal
-app
eari
ng
wh
ite
mat
ter
adja
cen
t to
th
e tu
mor
Mor
i et
al.35
2/0
3-di
men
sion
alD
ispl
acem
ent
of t
he
fibe
r tr
acts
in o
ne
pati
ent;
infi
ltra
tion
wit
hou
t di
spla
cem
ent
in
fibe
r tr
acki
ng
anot
her
pat
ien
tF
ocal
epi
leps
yK
rako
w e
t al
.841/
0FA
, Tr
Red
uce
d an
isot
ropy
, hig
h d
iffu
sivi
ty, a
nd
disp
lace
men
t of
mye
lin
ated
tra
cts
due
to a
m
alfo
rmat
ion
of c
orti
cal d
evel
opm
ent
was
det
ecte
d in
a p
atie
nt
wit
h fo
cal e
pile
psy
Ru
gg-G
un
n e
t al
.3730
/30
FA, T
rIn
crea
sed
diff
usi
vity
an
d re
duce
d an
isot
ropy
wer
e n
oted
wit
hin
th
e w
hit
e m
atte
r of
th
e le
ft t
empo
ral l
obe
in s
ubj
ects
wit
h e
lect
rocl
inic
al s
eizu
re o
nse
t lo
cali
zed
to t
he
left
te
mpo
ral l
obe
Eri
ksso
n e
t al
.8522
/30
FA, T
rIn
pat
ien
ts w
ith
a m
alfo
rmat
ion
of c
orti
cal d
evel
opm
ent,
red
uce
d an
isot
ropy
an
d in
crea
sed
diff
usi
vity
wer
e ob
serv
ed b
eyon
d th
e m
alfo
rmed
are
asM
SW
erri
ng
et a
l.866/
6FA
, Tr
In M
S p
atie
nts
th
e h
igh
est
diff
usi
vity
was
see
n in
des
tru
ctiv
e le
sion
s, w
her
eas
the
grea
test
ch
ange
in a
nis
otro
py w
as fo
un
d in
infl
amm
ator
y le
sion
sT
ievs
ky e
t al
.4112
/0FA
, RA
, EIn
acu
te le
sion
s, p
laqu
e ce
nte
rs h
ad h
igh
AD
C w
ith
red
uce
d an
isot
ropy
com
pare
d w
ith
ri
ms,
nor
mal
-app
eari
ng
wh
ite
mat
ter,
an
d ch
ron
ic le
sion
sN
usb
aum
et
al.87
13/1
2Tr
In M
S p
atie
nts
, mea
n w
hol
e-br
ain
dif
fusi
vity
was
ele
vate
dC
astr
iota
-20
/11
DD
iffu
sivi
ty w
as g
reat
er in
lesi
ons
of p
atie
nts
wit
h s
econ
dary
pro
gres
sive
MS
S
can
derb
eg e
t al
.43th
an in
th
ose
of p
atie
nts
wit
h r
elap
sin
g-re
mit
tin
g M
SB
amm
er e
t al
.8814
/9Tr
, FA
Dif
fusi
vity
was
sli
ghtl
y bu
t si
gnifi
can
tly
grea
ter
in n
orm
al-a
ppea
rin
g w
hit
e m
atte
r in
M
S p
atie
nts
th
an in
th
e w
hit
e m
atte
r of
con
trol
sF
ilip
pi e
t al
.4278
/20
FA, D
Nor
mal
-app
eari
ng
wh
ite
mat
ter
of M
S p
atie
nts
sh
owed
hig
her
dif
fusi
vity
an
d lo
wer
an
isot
ropy
th
an d
id w
hit
e m
atte
r of
con
trol
s
330 Kubicki et al.Harvard Rev Psychiatry
November/December 2002
Alz
hei
mer
’s
Ros
e et
al.57
11/9
LI
Pat
ien
ts s
how
ed lo
wer
an
isot
ropy
in t
he
asso
ciat
ion
wh
ite
mat
ter
fibe
r tr
acts
, su
ch a
s di
seas
eth
e sp
len
ium
of t
he
corp
us
call
osu
m, t
he
supe
rior
lon
gitu
din
al fa
scic
ulu
s, a
nd
the
cin
gulu
m, t
han
did
con
trol
sK
anta
rci e
t al
.5819
/55
AI†
AD
Cs
of t
he
hip
poca
mpu
s an
d th
e te
mpo
ral s
tem
, pos
teri
or c
ingu
late
, occ
ipit
al, a
nd
pari
etal
wh
ite
mat
ter
wer
e h
igh
er in
pat
ien
ts t
han
in c
ontr
ols
Sch
izop
hre
nia
Bu
chsb
aum
et
al.61
5/6
RA
Pat
ien
ts s
how
ed lo
wer
an
isot
ropy
in t
he
wh
ite
mat
ter
of t
he
pref
ron
tal c
orte
x th
an d
id
con
trol
sL
im e
t al
.6010
/10
FA, T
rA
nis
otro
py w
as lo
wer
in t
he
wh
ite
mat
ter
of p
atie
nts
th
an in
th
at o
f con
trol
sF
oon
g et
al.62
20/2
5FA
, Tr
Com
pare
d w
ith
con
trol
s, p
atie
nts
sh
owed
hig
her
dif
fusi
vity
an
d lo
wer
an
isot
ropy
in t
he
sple
niu
m b
ut
not
th
e ge
nu
of t
he
corp
us
call
osu
mA
gart
z et
al.63
20/2
4FA
, Tr
Pat
ien
ts s
how
ed lo
wer
an
isot
ropy
in t
he
sple
niu
m o
f th
e co
rpu
s ca
llos
um
as
wel
l as
a h
igh
er d
iffu
sivi
ty t
hro
ugh
out
the
enti
re v
olu
me
of w
hit
e m
atte
r co
mpa
red
to c
ontr
ols
Ste
el e
t al
.8910
/10
FAN
o di
ffer
ence
s in
an
isot
ropy
in p
refr
onta
l wh
ite
mat
ter
wer
e ob
serv
ed b
etw
een
con
trol
s an
d pa
tien
tsK
ubi
cki e
t al
.6415
/18
FAD
imin
ish
ed le
ft/r
igh
t as
ymm
etry
in a
nis
otro
py in
th
e u
nci
nat
e fa
scic
ulu
s w
as s
een
in
pati
ents
com
pare
d to
con
trol
sO
ther
con
diti
ons
affe
ctin
gw
hit
e m
atte
rH
IV
Pom
ara
et a
l.456/
9FA
, Tr,
PD
Fro
nta
l lob
e an
d in
tern
al c
apsu
le w
hit
e m
atte
r sh
owed
dec
reas
ed a
nis
otro
py in
pat
ien
ts
infe
ctio
nco
mpa
red
to c
ontr
ols
Fil
ippi
et
al.46
10/0
UA
, Tr
Dif
fusi
vity
dec
reas
ed in
th
e co
rpu
s ca
llos
um
an
d in
crea
sed
in t
he
subc
orti
cal f
ron
tal
and
pari
etal
reg
ions
in p
atie
nts
wit
h el
evat
ed v
iral
load
; no
chan
ges
in p
atie
nts
taki
ng
anti
retr
ovir
al d
rugs
Kra
bbe’
sG
uo
et a
l.448/
8R
AP
atie
nts
show
ed lo
wer
ani
sotr
opy
in w
hite
mat
ter
than
did
con
trol
s; fo
llow
ing
trea
men
t,
dise
ase
anis
otro
py in
crea
sed
but
was
sti
ll lo
wer
th
an in
con
trol
sC
hro
nic
Pfe
ffer
bau
m e
t al
.5015
/31
FA, I
CP
atie
nts
show
ed lo
wer
ani
sotr
opy
and
cohe
renc
e in
the
cor
pus
call
osum
tha
n di
d co
ntro
ls;
alco
hol
wor
kin
g m
emor
y an
d at
ten
tion
mea
sure
s co
rrel
ated
pos
itiv
ely
wit
h a
nis
otro
pyde
pen
den
ceA
LS
Ell
is e
t al
.4722
/20
FA, T
rA
nis
otro
py c
orre
late
d w
ith
dis
ease
sev
erit
y an
d u
pper
mot
or n
euro
n in
volv
emen
t;
diff
usi
vity
cor
rela
ted
wit
h d
isea
se d
ura
tion
Ulu
g et
al.49
4/0
UA
No
decr
ease
of a
nis
otro
py w
ith
in t
he
cort
icos
pin
al t
ract
was
det
ecta
ble
on M
RI
X-l
inke
d A
LD
Ito
et a
l.9011
/0A
DC
, FA
Aff
ecte
d ar
eas
of t
he
wh
ite
mat
ter
show
ed lo
wer
an
isot
ropy
th
an d
id u
naf
fect
ed a
reas
; fo
llow
-up
stud
ies
reve
aled
an
incr
ease
of t
he a
ffec
ted
area
on
the
FAm
aps,
att
ribu
tabl
e to
loss
of m
yeli
n-s
hea
th in
tegr
ity
CA
DA
SIL
Ch
abri
at e
t al.48
16/1
0V
R, T
rP
atie
nts
sh
owed
incr
ease
d di
ffu
sivi
ty a
nd
decr
ease
d an
isot
ropy
bot
h in
lesi
ons
and
in
norm
al-a
ppea
ring
whi
te m
atte
r co
mpa
red
to c
ontr
ols;
deg
ree
of a
bnor
mal
ity
corr
elat
edw
ith
cli
nic
al s
ever
ity
AD
C,a
ppar
ent
diff
usi
on c
oeffi
cien
t; A
I,an
isot
ropy
inde
x; A
LD
,adr
enol
euko
dyst
roph
y;A
LS
,am
yotr
oph
ic la
tera
l scl
eros
is; C
AD
AS
IL,c
ereb
ral a
uto
som
al d
omin
ant
arte
ri-
opat
hy w
ith
subc
orti
cal i
nfar
cts
and
leuk
oenc
epha
lopa
thy;
D,a
vera
ged
wat
er d
iffu
sion
coe
ffici
ent;
E,e
ccen
tric
ity;
FA,f
ract
iona
l ani
sotr
opy;
fMR
I,fu
ncti
onal
mag
neti
c re
sona
nce
imag
ing;
HIV
,hu
man
imm
un
odefi
cien
cy v
iru
s; I
C,i
nte
rvox
el c
oher
ence
; LI,
latt
ice
inde
x of
an
isot
ropy
; MR
I,m
agn
etic
res
onan
ce im
agin
g; M
S,m
ult
iple
scl
eros
is; P
D,p
roto
nde
nsi
ty;R
A,r
elat
ive
anis
otro
py; T
r, di
ffu
sion
tra
ce; U
A,u
ltim
ate
anis
otro
py; V
R,v
olu
me
rati
o.*A
ll fi
ndi
ngs
rep
orte
d ar
e si
gnifi
can
t at
p≤
0.05
.†A
regi
onal
mea
sure
of t
he
dire
ctio
nal
ity
of d
iffu
sion
.
TAB
LE
3.C
linic
al A
pp
licat
ion
s o
f Dif
fusi
on
Ten
sor
Imag
ing
(co
nt’d
.)
n (
pat
ien
ts/
Con
dit
ion
Stu
dy
con
trol
s)M
easu
re(s
)F
ind
ings
*
changes in myelinated structures, including the right opticradiation and the forceps major of the corpus callosum, whichcorresponded with clinically predictable symptoms. Thesecase studies suggest that DTI is a powerful new technologyfor investigating functional deficits caused by brain injury,as well as for predicting prognosis.
Brain tumors. DWI has not been particularly useful inbrain tumor differentiation, although it is helpful in detectingthe cystic components of tumors, early edema around thetumor, and ischemic lesions within the pathological mass. Incontrast, DTI can be implemented for modeling fiber-tractdisruptions or displacement caused by the tumor34,35 andcould be useful for early detection of spine metastases, aswell as for detection of corticospinal-tract disruptions ordisplacement.
Seizure disorders. Compared with its use in the detectionand differentiation of early stroke from brain tumor, DWI hasso far played only a minor role in routine interictal imaging.In rats, severe, acute focal damage (expressed in animals ascytotoxic edema, which causes a drop in diffusion) after pro-longed induced seizures can be detected with DWI, especiallywithin the amygdala and piriform cortex.35 These resultshave not been replicated in humans, presumably becausethey show a different mechanism of brain injury (vasogenicedema, which causes an increase in diffusion, coexists withthe cytotoxic edema).
A few existing DTI studies already demonstrate highersensitivity of DTI compared to structural MRI in detectingmalformations of cortical development. Such malformationsdisturb the orientation of the fiber tracts and are a commoncause of epilepsy. Rugg-Gunn and colleagues,37 for example,have shown that changes within the white matter of the lefttemporal lobe can be identified with DTI but not with struc-tural MRI.
Other Conditions Affecting White Matter. DWI has beenshown to be superior in detecting white matter abnormali-ties in MS—abnormalities that are not as readily observedin conventional structural MRI (“normal-appearing whitematter”).38–40 It is hoped that DTI will be of even greatervalue in detecting fiber-tract alterations due to demyelina-tion and axonal loss. DTI studies in patients with MS haveshown an increase in mean diffusivity and a decrease in dif-fusion anisotropy within acute lesions,41 as well as within thenormal-appearing white matter,42 most likely attributable toedema. In addition, DTI has been found to be useful in dif-ferentiating between two types of MS,43 secondary progres-sive and relapsing-remitting, which have different clinicalcourses.
Finally, DTI has demonstrated white matter fiber tractpathology in Krabbe’s disease,44 HIV infection,45,46 amy-
otrophic lateral sclerosis,47 cerebral autosomal dominant ar-teriopathy with subcortical infarcts and leukoencephalopa-thy,48 leukoencephalopathy,49 and chronic alcohol depend-ence.50 In such studies, DTI was able to detect and quantifytherapeutic responses to treatment that are believed to bethe result of decreased edema during the acute phase of in-flammation.44,46 The potential of DTI as an exploratory tool issuggested by a study of patients with chronic alcohol de-pendence50 that revealed a correlation between loss of work-ing memory and attention and a decrease in diffusionanisotropy within the corpus callosum.
Alzheimer’s disease. The early to moderate stages ofAlzheimer’s disease are characterized by impaired cognitionwith preserved mobility.51 Although the disease is believedmainly to affect gray matter, postmortem studies52,53 have re-vealed loss of axons and oligodendrocytes within the whitematter as well. Of note, a recent DWI study54 has shown re-duced diffusion within the splenium and body of the corpuscallosum, findings consistent with previous reports of atro-phy in the corpus callosum in patients with Alzheimer’s dis-ease.55,56 In addition, DTI studies conducted in the earlystages of the disease57,58 have revealed significant connectiv-ity disruptions within the association white matter fibertracts, including the temporal stem (uncinate fasciculus), cin-gulate fasciculus, corpus callosum, and superior longitudinalfasciculus, as well as the hippocampus. Thus, DTI studiesmay improve our ability to track progressive changes inAlzheimer’s disease and could possibly be used in the futureto evaluate changes in response to treatment.
Schizophrenia. Schizophrenia is a disorder of unknownetiology. Although many subtle brain abnormalities havebeen observed in this disorder (see review in reference 59),no brain lesions have been definitively correlated with manyof the functional deficits found in these patients. Of note, how-ever, are several DTI reports of decreased diffusion aniso-tropy in the white matter of persons with schizophrenia. Lossof orientation and organization of fiber tracts has beendetected in the whole white matter60 and in frontal whitematter,61 but also in particular fiber tracts such as the corpuscallosum62,63 and the uncinate fasciculus.64 Further DTIinvestigation of white matter fiber-tract abnormalities inschizophrenia may change how we view this disorder, partic-ularly in providing access to in vivo developmental studiesacross time.
Other possible uses in psychiatric disorders. To date,there are no studies using DTI to investigate white matterabnormalities that have been reported in other psychiatricconditions. MRI has revealed white matter hyperintensitiesin deep and periventricular white matter in patients with af-fective disorders;65 it has also shown deep white matter
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 331
lesions to be correlated with poor outcome in bipolar dis-order66 and with degree of residual dysfunction following asevere episode of depression.67 MRI studies in patientswith posttraumatic stress disorder have revealed nonspecificwhite matter lesions, as well as some functional deficits thatmight be attributed to the disconnection of specific corticalregions (i.e., the amygdala, Broca’s region, and the cingulatecortex).68 Correlations between psychiatric symptoms andwhite matter lesions could be further evaluated using DTI.Such testing might be able to determine the specificity of theparticular fiber tracts affected, as well as the extent of theirinvolvement.
Developmental StudiesDevelopmental DTI studies are only just beginning. Investi-gation of normal and abnormal brain development, however,should lead to a better understanding of brain maturation.DWI has, in fact, already revealed greater water diffusion inneonates than in adults,69,70 and there is evidence to suggestthat anisotropic diffusion is higher in full-term neonates thanin preterm neonates,3 a difference most likely due to themyelination of white matter fiber tracts. Such findings sug-gest that diffusion-imaging techniques can detect an increasein myelination during normal development.
In addition, anisotropic diffusion has been observed to de-cline as a result of age-related degenerative processes in-volving white matter fibers and myelin sheaths.71 Whetherthis change is due to normal aging or pathological aging re-mains to be determined. Table 4 summarizes all neurodevel-opmental studies utilizing DTI published through March2002.
Diffusion imaging offers the opportunity to evaluate bothnormal and pathological changes in white matter and brainconnectivity over the life span. Such changes may be impor-tant for understanding not only normal development but alsodifferences in cognitive abilities over time.
Future ApplicationsPotential future applications of DTI include visualization ofthe anatomical connections among different parts of thebrain. Diffusion tensor tractography (see Figure 3), proposedby several authors,72–75 uses the principal diffusion directionmeasured with DTI to compute the pathways of completenerve fiber tracts. The tracing is performed by first definingregions of interest. Then, starting from points (“seed points”)selected within this region and following the spatially inter-polated direction of maximum diffusion in neighboring voxels,the path of fibers within a fiber tract is defined. Such tracking,which is done repetitively with multiple seed points, createsa contiguous path that defines the fiber tract of interest.
Visualization is then performed in three dimensions to de-pict the white matter fiber tract. Progressing from an exam-ination of anisotropy to a more elaborate analysis of the re-lationship between neighboring diffusion ellipsoids opens thepossibility for assessing, in vivo, axonal fiber connectivity andfunctional links among brain regions.
A slightly different approach to tensor tractography, de-scribed by Westin and colleagues,14 attempts to reduce prob-lems encountered when tracing fibers in complex regions(such as where fiber tracts merge, branch, or cross within avoxel) by defining the direction of the trace path in a novelway. Rather than following the direction of the maximum dif-fusivity (the direction of the major axis of the diffusion ellip-soid), this approach “bends” the trace with a strong bias to-ward this direction. Another approach regularizes the tracepath based on its curvature—that is, fixes the curve accordingto a predefined parameter.
Diffusion tensor tractography, combined with informationfrom conventional and functional MR imaging, can provide apowerful tool for neurosurgical planning, especially when sur-gery occurs in the vicinity of vital nerve fiber tracts. Tracingand mapping the passage of functionally relevant fiber tractsalong the tumors is as important as mapping cortical func-
332 Kubicki et al.Harvard Rev Psychiatry
November/December 2002
TABLE 4. Neurodevelopmental Studies Utilizing DTI
Study n Measure(s) Findings*
Hüppi et al.3 24 infants: 17 preterm, Diffusivity, RA Anisotropy was higher the closer birth was to term7 full-term
Klingberg et al.91 7 children, 5 adults FA, Tr Anisotropy in frontal white matter was lower in children than in adults
Pfefferbaum et al.71 31 healthy men, ages Diffusivity, FA, IC Anisotropy declined with age in all regions except the splenium 23–76 y of the corpus callosum
Nusbaum et al.92 20 healthy volunteers, RA, diffusivity Anisotropy decreased significantly with increasing age in ages 20–91 y histograms periventricular white matter, frontal white matter, and the
genu and splenium of the corpus callosum
FA, fractional anisotropy; IC, intervoxel coherence; RA, relative anisotropy; Tr, diffusion trace.*All findings reported are significant at p ≤ 0.05.
tions adjacent to tumors. The information gathered withthese complementary techniques helps the neurosurgeon todecide where tumor tissue can be excised without permanentneurological consequences.
In addition, DTI, as mentioned above, can be used to followup surgical and neurological treatment (by assessing the re-generation and/or remyelination of the affected fiber tracts),as well as to monitor the effects of medication. Finally, in dis-orders such as schizophrenia, where gross brain abnormali-ties are not evident, DTI may offer an opportunity to evaluatesubtle changes in white matter fiber tracts that are relatedto neurocognitive abnormalities observed in this disorder.Such information might further our knowledge of brain-connectivity abnormalities and lead to more-targeted phar-macological treatments as well as to a better understandingof brain-behavior links in this devastating disorder.
In summary, DTI has opened up new research possibili-ties in areas that previously relied largely upon postmortemstudies. For the first time, the intricate connective architec-ture of the most complex human organ can be studied nonin-vasively. Other potentially important applications for thistechnique, such as characterization of cardiac muscle tissuearchitecture, diagnosis of liver disease, mapping of tissuetemperature, and diffusion spectroscopy, are beyond the scopeof this article. It is clear that DTI could revolutionize what isknown in many different domains of medicine and disease.The ability to visualize white matter fiber tracts in the hu-man brain, in vivo, will likely be critical to a new under-standing of brain structure and function, both in normal in-dividuals and in those with a neuropsychiatric disorder.
The authors would like to thank Marie Fairbanks for her admin-istrative assistance.
REFERENCES
1. Beaulieu C, Allen PS. Determinants of anisotropic water diffu-sion in nerves. Magn Reson Med 1994;31:394–400.
2. Wimberger DM, Roberts TP, Barkovich AJ, Prayer LM, MoseleyME, Kucharczyk J. Identification of “premyelination” bydiffusion-weighted MRI. J Comput Assist Tomogr 1995;19:28–33.
3. Hüppi PS, Maier SE, Peled S, Zientara GP, Barnes PD, JoleszFA, et al. Microstructural development of human newborn cere-bral white matter assessed in vivo by diffusion tensor magneticresonance imaging. Pediatr Res 1998;44:584–90.
4. Le Bihan D, Breton E, Lallemand D, Grenier P, Cabanis E,Laval-Jeantet M. MR imaging of intravoxel incoherent motions:application to diffusion and perfusion in neurologic disorders.Radiology 1986;161:401–7.
5. Pierpaoli C, Jezzard P, Basser PJ, Barnett A, Di Chiro G. Diffu-sion tensor MR imaging of the human brain. Radiology 1996;201:637–48.
6. Basser PJ, Mattiello J, Le Bihan D. MR diffusion tensor spec-troscopy and imaging. Biophys J 1994;66:259–67.
7. Turner R, Le Bihan D, Maier J, Vavrek R, Hedges LK, Pekar J.Echo-planar imaging of intravoxel incoherent motion. Radiology1990;177:407–14.
8. Ordidge RJ, Helpern JA, Qing ZX, Knight RA, Nagesh V.Correction of motional artifacts in diffusion-weighted MR images using navigator echoes. Magn Reson Imaging 1994;12:455–60.
9. Trouard TP, Theilmann RJ, Altbach MI, Gmitro AF. High-resolution diffusion imaging with DIFRAD-FSE (diffusion-weighted radial acquisition of data with fast spin-echo) MRI.Magn Reson Med 1999;42:11–8.
10. Maier SE, Gudbjartsson H, Patz S, Hsu L, Lovblad KO, EdelmanRR, et al. Line scan diffusion imaging: characterization inhealthy subjects and stroke patients. Am J Roentgenol 1998;171:85–93.
11. Maier SE. Slab scan diffusion imaging. Magn Reson Med2001;46:1136–43.
12. Westin CF, Maier SE, Mamata H, Nabavi A, Jolesz FA, Kikinis R.Processing and visualization for diffusion tensor MRI. Med Im-age Anal 2002;6:93–108.
13. Basser PJ, Pierpaoli C. Microstructural and physiological fea-tures of tissues elucidated by quantitative-diffusion-tensor MRI.J Magn Reson B 1996;111:209–19.
14. Reese TG, Weisskoff RM, Smith RN, Rosen BR, Dinsmore RE,Wedeen VJ. Imaging myocardial fiber architecture in vivo withmagnetic resonance. Magn Reson Med 1995;34:786–91.
15. Makris N, Worth AJ, Sorensen AG, Papadimitriou GM, Wu O,Reese TG, et al. Morphometry of in vivo human white matter as-sociation pathways with diffusion-weighted magnetic resonanceimaging. Ann Neurol 1997;42:951–62.
16. Pierpaoli C, Basser PJ. Toward a quantitative assessment of dif-fusion anisotropy. Magn Reson Med 1996;36:893–906.
17. Peled S, Gudbjartsson H, Westin CF, Kikinis R, Jolesz FA. Mag-netic resonance imaging shows orientation and asymmetry ofwhite matter fiber tracts. Brain Res 1998;780:27–33.
18. Moseley ME, Cohen Y, Mintorovitch J, Chileuitt L, Shimizu H,Kucharczyk J, et al. Early detection of regional cerebral ischemiain cats: comparison of diffusion- and T2-weighted MRI and spec-troscopy. Magn Reson Med 1990;14:330–46.
19. Warach S, Chien D, Li W, Ronthal M, Edelman RR. Fast mag-netic resonance diffusion-weighted imaging of acute humanstroke. Neurology 1992;42:1717–23.
20. Warach S, Gaa J, Siewert B, Wielopolski P, Edelman RR. Acutehuman stroke studied by whole brain echo planar diffusion-weighted magnetic resonance imaging. Ann Neurol 1995;37:231–41.
21. Tsuruda JS, Chew WM, Moseley ME, Norman D. Diffusion-weighted MR imaging of the brain: value of differentiating be-tween extraaxial cysts and epidermoid tumors. Am J Roentgenol1990;155:1059–65.
22. Hajnal JV, Doran M, Hall AS, Collins AG, Oatridge A, PennockJM, et al. MR imaging of anisotropically restricted diffusion ofwater in the nervous system: technical, anatomic, and pathologicconsiderations. J Comput Assist Tomogr 1991;15:1–18.
23. Chien D, Kwong KK, Gress DR, Buonanno FS, Buxton RB, Rosen
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 333
BR. MR diffusion imaging of cerebral infarction in humans. AmJ Neuroradiol 1992;13:1097–102.
24. Warach S, Dashe JF, Edelman RR. Clinical outcome in ischemicstroke predicted by early diffusion-weighted and perfusion mag-netic resonance imaging: a preliminary analysis. J Cereb BloodFlow Metab 1996;16:53–9.
25. Rother J, De Crespigny AJ, D’Arceuil H, Iwai K, Moseley ME.Recovery of apparent diffusion coefficient after ischemia-inducedspreading depression relates to cerebral perfusion gradient.Stroke 1996;27:980–6.
26. Zelaya F, Flood N, Chalk JB, Wang D, Doddrell DM, StrugnellW, et al. An evaluation of the time dependence of the anisotropyof the water diffusion tensor in acute human ischemia. Magn Re-son Imaging 1999;17:331–48.
27. Werring DJ, Toosy AT, Clark CA, Parker GJ, Barker GJ, MillerDH, et al. Diffusion tensor imaging can detect and quantify cor-ticospinal tract degeneration after stroke. J Neurol NeurosurgPsychiatry 2000;69:269–72.
28. Pierpaoli C, Barnett A, Pajevic S, Chen R, Penix LR, Virta A,et al. Water diffusion changes in wallerian degeneration andtheir dependence on white matter architecture. Neuroimage2001;13:1174–85.
29. Mukherjee P, Bahn MM, McKinstry RC, Shimony JS, Cull TS,Akbudak E, et al. Differences between gray matter and whitematter water diffusion in stroke: diffusion-tensor MR imagingin 12 patients. Radiology 2000;215:211–20.
30. Ito J, Marmarou A, Barzo P, Fatouros P, Corwin F. Characteri-zation of edema by diffusion-weighted imaging in experimentaltraumatic brain injury. J Neurosurg 1996;84:97–103.
31. Barzo P, Marmarou A, Fatouros P, Hayasaki K, Corwin F. Con-tribution of vasogenic and cellular edema to traumatic brainswelling measured by diffusion-weighted imaging. J Neurosurg1997;87:900–7.
32. Werring DJ, Clark CA, Barker GJ, Miller DH, Parker GJ, Bram-mer MJ, et al. The structural and functional mechanisms ofmotor recovery: complementary use of diffusion tensor and func-tional magnetic resonance imaging in a traumatic injury ofthe internal capsule. J Neurol Neurosurg Psychiatry 1998;65:863–9.
33. Wieshmann UC, Symms MR, Clark CA, Lemieux L, Parker GJ,Barker GJ, et al. Blunt-head trauma associated with widespreadwater-diffusion changes [Letter]. Lancet 1999;353:1242–3.
34. Wieshmann UC, Symms MR, Parker GJ, Clark CA, Lemieux L,Barker GJ, et al. Diffusion tensor imaging demonstrates de-viation of fibres in normal appearing white matter adjacentto a brain tumour. J Neurol Neurosurg Psychiatry 2000;68:501–3.
35. Mori S, Frederiksen K, Van Zijl PC, Stieltjes B, Kraut MA,Solaiyappan M, et al. Brain white matter anatomy of tumorpatients evaluated with diffusion tensor imaging. Ann Neurol2002;51:377–80.
36. Lynch MW, Rutecki PA, Sutula TP. The effects of seizures on thebrain. Curr Opin Neurol 1996;9:97–102.
37. Rugg-Gunn FJ, Eriksson SH, Symms MR, Barker GJ, DuncanJS. Diffusion tensor imaging of cryptogenic and acquired partialepilepsies. Brain 2001;124:627–36.
38. Christiansen P, Gideon P, Thomsen C, Stubgaard M, HenriksenO, Larsson HB. Increased water self-diffusion in chronic plaques
and in apparently normal white matter in patients with multiplesclerosis. Acta Neurol Scand 1993;87:195–9.
39. Droogan AG, Clark CA, Werring DJ, Barker GJ, McDonald WI,Miller DH. Comparison of multiple sclerosis clinical subgroupsusing navigated spin echo diffusion-weighted imaging. Magn Re-son Imaging 1999;17:653–61.
40. Cercignani M, Iannucci G, Rocca MA, Comi G, Horsfield MA, Fil-ippi M. Pathologic damage in MS assessed by diffusion-weightedand magnetization transfer MRI. Neurology 2000;54:1139–44.
41. Tievsky AL, Ptak T, Farkas J. Investigation of apparent diffu-sion coefficient and diffusion tensor anisotropy in acute andchronic multiple sclerosis lesions. Am J Neuroradiol 1999;20:1491–9.
42. Filippi M, Cercignani M, Inglese M, Horsfield MA, Comi G. Dif-fusion tensor magnetic resonance imaging in multiple sclerosis.Neurology 2001;56:304–11.
43. Castriota-Scanderbeg A, Tomaiuolo F, Sabatini U, Nocentini R,Grasso MG, Caltagirone C. Demyelinating plaques in relapsing-remitting and secondary-progressive multiple sclerosis: assess-ment with diffusion MR imaging. Am J Neuroradiol 2000;21:862–8.
44. Guo AC, Petrella JR, Kurtzberg J, Provenzale JM. Evaluation ofwhite matter anisotropy in Krabbe disease with diffusion tensorMR imaging: initial experience. Radiology 2001;218:809–15.
45. Pomara N, Crandall DT, Choi SJ, Johnson G, Lim KO. Whitematter abnormalities in HIV-1 infection: a diffusion tensor im-aging study. Psychiatry Res 2001;106:15–24.
46. Filippi CG, Ulug AM, Ryan E, Ferrando SJ, Van Gorp W. Diffu-sion tensor imaging of patients with HIV and normal-appearingwhite matter on MR images of the brain. Am J Neuroradiol2001;22:277–83.
47. Ellis CM, Simmons A, Jones DK, Bland J, Dawson JM, HorsfieldMA, et al. Diffusion tensor MRI assesses corticospinal tract dam-age in ALS. Neurology 1999;53:1051–8.
48. Chabriat H, Pappata S, Poupon C, Clark CA, Vahedi K, PouponF, et al. Clinical severity in CADASIL related to ultrastructuraldamage in white matter: in vivo study with diffusion tensor MRI.Stroke 1999;30:2637–43.
49. Ulug AM, Moore DF, Bojko AS, Zimmerman RD. Clinical use ofdiffusion-tensor imaging for diseases causing neuronal and ax-onal damage. Am J Neuroradiol 1999;20:1044–8.
50. Pfefferbaum A, Sullivan EV, Hedehus M, Adalsteinsson E, LimKO, Moseley M. In vivo detection and functional correlates ofwhite matter microstructural disruption in chronic alcoholism.Alcohol Clin Exp Res 2000;24:1214–21.
51. Goldman WP, Baty JD, Buckles VD, Sahrmann S, Morris JC.Motor dysfunction in mildly demented AD individuals withoutextrapyramidal signs. Neurology 1999;53:956–62.
52. Brun A, Englund E. A white matter disorder in dementia of theAlzheimer type: a pathoanatomical study. Ann Neurol 1986;19:253–62.
53. Englund E. Neuropathology of white matter changes inAlzheimer’s disease and vascular dementia. Dement GeriatrCogn Disord 1998;9(suppl 1):6–12.
54. Hanyu H, Asano T, Sakurai H, Imon Y, Iwamoto T, Takasaki M,et al. Diffusion-weighted and magnetization transfer imaging ofthe corpus callosum in Alzheimer’s disease. J Neurol Sci1999;167:37–44.
334 Kubicki et al.Harvard Rev Psychiatry
November/December 2002
55. Weis S, Jellinger K, Wenger E. Morphometry of the corpus callo-sum in normal aging and Alzheimer’s disease. J Neural TransmSuppl 1991;33:35–8.
56. Janowsky JS, Kaye JA, Carper RA. Atrophy of the corpus callo-sum in Alzheimer’s disease versus healthy aging. J Am GeriatrSoc 1996;44:798–803.
57. Rose SE, Chen F, Chalk JB, Zelaya FO, Strugnell WE, BensonM, et al. Loss of connectivity in Alzheimer’s disease: an evalua-tion of white matter tract integrity with colour coded MR dif-fusion tensor imaging. J Neurol Neurosurg Psychiatry 2000;69:528–30.
58. Kantarci K, Jack CR Jr, Xu YC, Campeau NG, O’Brien PC, SmithGE, et al. Mild cognitive impairment and Alzheimer disease: re-gional diffusivity of water. Radiology 2001;219:101–7.
59. Shenton ME, Dickey CC, Frumin M, McCarley RW. A review ofMRI findings in schizophrenia. Schizophr Res 2001;49:1–52.
60. Lim KO, Hedehus M, Moseley M, De Crespigny A, Sullivan EV,Pfefferbaum A. Compromised white matter tract integrity inschizophrenia inferred from diffusion tensor imaging. Arch GenPsychiatry 1999;56:367–74.
61. Buchsbaum MS, Tang CY, Peled S, Gudbjartsson H, Lu D, Haz-lett EA, et al. MRI white matter diffusion anisotropy and PETmetabolic rate in schizophrenia. Neuroreport 1998;9:425–30.
62. Foong J, Maier M, Clark CA, Barker GJ, Miller DH, Ron MA.Neuropathological abnormalities of the corpus callosum in schiz-ophrenia: a diffusion tensor imaging study. J Neurol NeurosurgPsychiatry 2000;68:242–4.
63. Agartz I, Andersson JL, Skare S. Abnormal brain white matter inschizophrenia: a diffusion tensor imaging study. Neuroreport2001;12:2251–4.
64. Kubicki M, Westin CF, Maier S, Frumin M, Nestor PG, SalisburyDF, et al. Uncinate fasciculus findings in schizophrenia: a mag-netic resonance diffusion tensor imaging study. Am J Psychiatry2002;159:813–20.
65. Altshuler LL, Curran JG, Hauser P, Mintz J, Denicoff K, Post R.T2 hyperintensities in bipolar disorder: magnetic resonance im-aging comparison and literature meta-analysis. Am J Psychia-try 1995;152:1139–44.
66. Moore PB, Shepherd DJ, Eccleston D, Macmillan IC, GoswamiU, McAllister VL, et al. Cerebral white matter lesions in bipolaraffective disorder: relationship to outcome. Br J Psychiatry2001;178:172–6.
67. Hickie I, Scott E, Wilhelm K, Brodaty H. Subcortical hyperin-tensities on magnetic resonance imaging in patients with severedepression—a longitudinal evaluation. Biol Psychiatry 1997;42:367–74.
68. Pitman RK, Shin LM, Rauch SL. Investigating the pathogene-sis of posttraumatic stress disorder with neuroimaging. J ClinPsychiatry 2001;62(suppl 17):47–54.
69. Sakuma H, Nomura Y, Takeda K, Tagami T, Nakagawa T, Tam-agawa Y, et al. Adult and neonatal human brain: diffusionalanisotropy and myelination with diffusion-weighted MR imag-ing. Radiology 1991;180:229–33.
70. Neil JJ, Shiran SI, McKinstry RC, Schefft GL, Snyder AZ, AlmliCR, et al. Normal brain in human newborns: apparent diffusioncoefficient and diffusion anisotropy measured by using diffusiontensor MR imaging. Radiology 1998;209:57–66.
71. Pfefferbaum A, Sullivan EV, Hedehus M, Lim KO, Adalsteins-
son E, Moseley M. Age-related decline in brain white matteranisotropy measured with spatially corrected echo-planar diffu-sion tensor imaging. Magn Reson Med 2000;44:259–68.
72. Poupon C, Clark CA, Frouin V, Régis J, Bloch I, Le Bihan D, et al.Regularization of diffusion-based direction maps for the track-ing of brain white matter fascicles. Neuroimage 2000;12:184–95.
73. Poupon C, Mangin J, Clark CA, Frouin V, Régis J, Le Bihan D, etal. Towards inference of human brain connectivity from MR dif-fusion tensor data. Med Image Anal 2001;5:1–15.
74. Wiegell MR, Larsson HB, Wedeen VJ. Fiber crossing in humanbrain depicted with diffusion tensor MR imaging. Radiology2000;217:897–903.
75. Basser PJ, Pajevic S, Pierpaoli C, Duda J, Aldroubi A. In vivofiber tractography using DT-MRI data. Magn Reson Med 2000;44:625–32.
76. Ulug AM, Van Zihl PC. Orientation-independent diffusion im-aging without tensor diagonalization: anisotropy definitionsbased on physical attributes of the diffusion ellipsoid. J MagnReson Imaging 1999;9:804–13.
77. Shimony JS, McKinstry RC, Akbudak E, Aronovitz JA, SnyderAZ, Lori NF, et al. Quantitative diffusion-tensor anisotropy brainMR imaging: normative human data and anatomic analysis. Ra-diology 1999;212:770–84.
78. Westin CF, Peled S, Gudbjartsson H, Kikinis R, Jolesz FA. Geo-metrical diffusion measures from tensor basis analysis [Ab-stract]. Proceedings of the Fifth Scientific Meeting of the Inter-national Society of Magnetic Resonance in Medicine, Vancouver,British Columbia, April 1997:1742.
79. Armitage PA, Bastin ME. Selecting an appropriate anisotropyindex for displaying diffusion tensor imaging data with improvedcontrast and sensitivity. Magn Reson Med 2000;44:117–21.
80. Jones DK, Lythgoe D, Horsfield MA, Simmons A, Williams SC,Markus HS. Characterization of white matter damage in is-chemic leukoaraiosis with diffusion tensor MRI. Stroke 1999;30:393–7.
81. Sorensen AG, Wu O, Copen WA, Davis TL, Gonzalez RG, Ko-roshetz WJ, et al. Human acute cerebral ischemia: detection ofchanges in water diffusion anisotropy by using MR imaging. Ra-diology 1999;212:785–92.
82. Jones DK, Dardis R, Ervine M, Horsfield MA, Jeffree M, Sim-mons A, et al. Cluster analysis of diffusion tensor magneticresonance images in human head injury. Neurosurgery 2000;47:306–13.
83. Bastin ME, Delgado M, Whittle IR, Cannon J, Wardlaw JM.The use of diffusion tensor imaging in quantifying the effectof dexamethasone on brain tumours. Neuroreport 1999;10:1385–91.
84. Krakow K, Wieshmann UC, Woermann FG, Symms MR, McLeanMA, Lemieux L, et al. Multimodal MR imaging: functional, dif-fusion tensor, and chemical shift imaging in a patient withlocalization-related epilepsy. Epilepsia 1999;40:1459–62.
85. Eriksson SH, Rugg-Gunn FJ, Symms MR, Barker GJ, DuncanJS. Diffusion tensor imaging in patients with epilepsy and mal-formations of cortical development. Brain 2001;124:617–26.
86. Werring DJ, Clark CA, Barker GJ, Thompson AJ, Miller DH. Dif-fusion tensor imaging of lesions and normal-appearing whitematter in multiple sclerosis. Neurology 1999;52:1626–32.
87. Nusbaum AO, Tang CY, Wei T, Buchsbaum MS, Atlas SW. Whole-
Harvard Rev Psychiatry
Volume 10, Number 6 Kubicki et al. 335
brain diffusion MR histograms differ between MS subtypes. Neu-rology 2000;54:1421–7.
88. Bammer R, Augustin M, Strasser-Fuchs S, Seifert T, Kapeller P,Stollberger R, et al. Magnetic resonance diffusion tensor imagingfor characterizing diffuse and focal white matter abnormalities inmultiple sclerosis. Magn Reson Med 2000;44:583–91.
89. Steel RM, Bastin ME, McConnell S, Marshall I, Cunningham-Owens DG, Lawrie SM, et al. Diffusion tensor imaging (DTI) andproton magnetic resonance spectroscopy (1H MRS) in schizo-phrenic subjects and normal controls. Psychiatry Res 2001;106:161–70.
90. Ito R, Melhem ER, Mori S, Eichler FS, Raymond GV, MoserHW. Diffusion tensor brain MR imaging in X-linked cerebraladrenoleukodystrophy. Neurology 2001;56:544–7.
91. Klingberg T, Vaidya CJ, Gabrieli JD, Moseley ME, Hedehus M.Myelination and organization of the frontal white matter inchildren: a diffusion tensor MRI study. Neuroreport 1999;10:2817–21.
92. Nusbaum AO, Tang CY, Buchsbaum MS, Wei TC, Atlas SW. Re-gional and global changes in cerebral diffusion with normal ag-ing. Am J Neuroradiol 2001;22:136–42.
336 Kubicki et al.Harvard Rev Psychiatry
November/December 2002