Diffusion Tensor Imaging of Mild to Moderate Blast-Related
Traumatic Brain Injury and Its Sequelae
Harvey S. Levin,1,2Elisabeth Wilde,1,2,3Maya Troyanskaya,1Nancy J. Petersen,2,4Randall Scheibel,1,2
Mary Newsome,1,2Majdi Radaideh,2,3Trevor Wu,5Ragini Yallampalli,1Zili Chu,3and Xiaoqi Li1
To evaluate the effects of mild to moderate blast-related traumatic brain injury (TBI) on the microstructure of
brain white matter (WM) and neurobehavioral outcomes, we studied 37 veterans and service members (mean
age 31.5 years, SD¼7.2; post-injury interval 871.5 days; SD¼343.1), whose report of acute neurological status
was consistent with sustaining mild to moderate TBI due to blast while serving in Iraq or Afghanistan. Fifteen
veterans without a history of TBI or exposure to blast (mean age 31.4 years, SD¼5.4) served as a comparison
group, including seven subjects with extracranial injury (post-injury interval 919.5 days, SD¼455.1), and eight
who were uninjured. Magnetic resonance imaging disclosed focal lesions in five TBI participants. Post-
concussion symptoms (Neurobehavioral Symptom Inventory), post-traumatic stress disorder (PTSD) symptoms
(PTSD Checklist-Civilian), and global distress and depression (Brief Symptom Inventory) were worse in the TBI
participants than the comparison group, but no group differences were found in perceived physical or mental
functioning (SF-12). Verbal memory (Selective Reminding) was less efficient in the TBI group, but there were no
group differences in nonverbal memory (Selective Reminding) or decision making (Iowa Gambling Task). Verbal
memory in the TBI group was unrelated to PTSD severity. Diffusion tensor imaging (DTI) using tractography,
standard single-slice region-of-interest measurement, and voxel-based analysis disclosed no group differences in
fractional anisotropy (FA) and apparent diffusion coefficient (ADC). However, FA of the left and right posterior
internal capsule and left corticospinal tract was positively correlated with total words consistently recalled,
whereas ADC for the left and right uncinate fasciculi and left posterior internal capsule was negatively corre-
lated with this measure of verbal memory. Correlations of DTI variables with symptom measures were non-
significant and inconsistent. Our data do not show WM injury in mild to moderate blast-related TBI in veterans
despite their residual symptoms and difficulty in verbal memory. Limitations of the study and implications for
future research are also discussed.
Key words: blast-related traumatic brain injury; diffusion tensor imaging; outcome
Iraq and Afghanistan (Warden, 2006), accounting for one-
fourth of medical evacuations (French and Parkinson, 2008;
Galarneau et al., 2008). An explosive mechanism was linked
to 78% of the clinician-confirmed injuries sustained by U.S.
service members over a 3-year period, and the head and neck
were involved in 30% of these injuries (Owens et al., 2008). In
post-deployment questionnaires, 12% (Schneiderman et al.,
2008) to 23% (Hoge et al., 2008; Terrio et al., 2009) of returnees
raumatic brain injury (TBI) is a leading cause of
morbidity and disability in service members deployed in
88% of these injuries (Hoge et al., 2008; Terrio et al., 2009), and
98% had acute injury features consistent with mild TBI (Hoge
et al.,2008). Of1292 soldiers in anArmy unit who reported on
a post-deployment questionnaire that they had sustained an
injury approximately 6 months earlier, a clinical interview
confirmed that TBI associated with brief loss or alteration of
consciousness had occurred in 907 (22.8%) of these service
members (Terrio et al., 2009). Despite the apparent mild na-
ture of clinician-confirmed TBI studied by Terrio and associ-
ates (2009), post-concussion symptoms reported at 6 months
post-injury included memory deficit (16.3%), headache
Departments of1Physical Medicine and Rehabilitation,3Radiology, and4Medicine, Baylor College of Medicine, Houston, Texas.
2Michael E. DeBakey VA Medical Center, Houston, Texas.
5Department of Psychology, Brigham Young University, Provo, Utah.
JOURNAL OF NEUROTRAUMA 27:683–694 (April 2010)
ª Mary Ann Liebert, Inc.
(20.20%), and irritability (21.30%). Other studies have con-
firmed the presence of cognitive, behavioral, and emotional
post-concussion symptoms on post-deployment question-
naires completed by service members who reported having
sustained injuries consistent with mild TBI (Belanger et al.,
2009; Hoge et al., 2008; Schneiderman et al., 2008). Comorbid
post-traumatic stress disorder (PTSD) and depression com-
plicate outcomes in nearly half of returnees reporting mild
blast-related TBI(Hogeet al.,2008;Schneiderman etal.,2008),
posing a challenge for treatment. However, the neural
mechanisms mediating these symptoms that are present
months after mild blast-related TBI are unknown, and imag-
ing studies of those affected are sparse.
The blast wave of increased pressure produced by an
explosion may be construed as the primary mechanism of
blast-related TBI (Cernak and Noble-Haeusslein, 2009), but
secondary and tertiary injuries that occur when the body is
struck by projectiles such as debris or propelled by the explo-
sion reportedly account for the majority of injuries resulting
from improvised explosive devices (Champion et al., 2009).
Although there is a lack of consensus concerning the contri-
bution of specific neural mechanisms to brain injury resulting
from blast (Cernak and Noble-Haeusslein, 2009), a recent re-
port showed that the cognitive functioning in service mem-
bers who had sustained TBI was similar across blast, motor
vehicle crashes, and non-blast blunt injury (Belanger et al.,
2009). This similarity in the pattern of cognitive findings has
given rise to the possibility that blast-related TBI may involve
axonal injury similar to the neuropathology that has been
reported in civilian TBI (Blumbergs et al., 1994;Oppenheimer,
1968). Disruption of white matter (WM) integrity with func-
in slowing of cognitive processing and other cognitive im-
pairments related to civilian TBI severity (Meythaler et al.,
2001; Povlishock and Katz, 2005). The lack of brain imaging
studies of service members and veterans with blast-related
TBI has limited progress in characterizing the pathology of
these injuries, giving impetus to performing this study.
To evaluate neuropathology in a sample of Operation En-
during Freedom and Operation Iraqi Freedom (OEF=OIF)
veterans who reported blast-induced TBI, we utilized mag-
netic resonance imaging (MRI) with diffusion tensor imaging
(DTI), a relatively new MRI technique that evaluates WM
microstructure by measuring diffusion of water. WM tracts
normally constrain the isotropic diffusion of water, a charac-
teristic indexed by fractional anisotropy (FA), wherein an FA
value approaching 1.0 reflects maximal anisotropic diffusion,
with values approaching zero indicating compromise of WM
integrity. Conversely, the apparent diffusion coefficient
(ADC) generally increases with TBI severity, especially when
DTI is performed in the chronic post-injury period. DTI is
sensitive to a continuum of civilian TBI severity (Benson et al.,
2007; Kraus et al., 2007) and chronicity (Bazarian et al., 2007;
Inglese et al., 2005; Kraus et al., 2007; Wilde et al., 2008). In
contrast to reports of lower FA in chronic mild TBI patients
(Kraus et al., 2007), DTI within the first week (Bazarian et al.,
2007; Chu et al., 2010; Wilde et al., 2008; Wu et al., 2009) after
mild TBI showed increased FA and lower ADC consistent
with cytotoxic edema, a potentially reversible condition. The
relationship of DTI variables to cognitive measures has been
found to be different in civilians with chronic mild TBI com-
pared with uninjured subjects, but between-group differences
in FA have not been consistent across brain regions (Niogi
et al., 2008). Some issues in the interpretation of these DTI
studies have been inclusion of patients with structural lesions
(Benson et al., 2007; Inglese et al., 2005; Niogi et al., 2008), and
unavailability of acute imaging findings (Kraus et al., 2007).
We evaluated the integrity of brain WM in veterans and
service members who screened positive for mild to moderate
DTI would be positively related to injury severity and post-
concussion symptoms, but negatively related to cognitive
functioning and health-related quality of life (HRQOL). With
the high rate of comorbid PTSD and depression in OEF=OIF
veterans (Hoge et al., 2008; Invisible Wounds of War, 2008;
Schneiderman et al., 2008; Seal et al., 2009) with blast-related
mild TBI, we also analyzed these conditions and general
This prospective, dual-cohort design recruited two groups
of post-deployment OEF=OIF veterans and service members,
including a group with self-report of mild to moderate TBI
due to blast, and a comparison group without blast exposure
who sustained injury to other body regions or had no injury.
This comparison group controlled for the adverse effects of
deployment-related stress on sustained attention, memory,
and negative affect (Vasterling et al., 2006).
The study was approved by the Institutional Review Board
and the Michael E. DeBakey Veterans Affairs Medical Center
(MEDVAMC) Research and Development Board, and par-
ticipants provided written informed consent. Recruitment
was done through physician referral, research database
search at the MEDVAMC, and through advertisement in the
newspaper of a regional Army post. Eligibility criteria for the
TBI group included diagnosis of blast-related TBI by a phy-
sician, and self-reported injury features consistent with TBI
due to blast (i.e., one or more blast injuries associated with or
a period of confusion, and=or post-traumatic amnesia (PTA)
immediately after the blast (French and Parkinson, 2008). As
part of the clinical assessment of these patients, an interview
with a physician identified loss or alteration of consciousness
at the time of injury associated with blast. This interview also
queried the patient regarding loss of memory for events prior
to and following the injury. Other eligibility criteria in both
groups were age 18–54 years; within 42 months post-injury
(for all injured subjects); right-handedness; no history of se-
vere neurological disorders, schizophrenia, or bipolar disor-
der; no pre-deployment hospitalization for TBI; no current
substance abuse; and no contraindications to undergoing
MRI. All participants were screened for previous and current
substance abuse (Maisto et al., 2000; Saunders et al., 1993;
Skinner, 1982). Of 236 veterans and service members screened
for this study, 182 individuals were excluded: 60 had post-
injury intervals longer than 42 months, 42 declined or lacked
contact information, 30 reported exposure to blasts, but
without altered consciousness, 21 had history of TBI, but not
due to blast, 15 had alcohol or=and drug abuse problems,
684 LEVIN ET AL.
9 were diagnosed with severe neurological=or psychiatric
disorders, and 5 could not participate due to claustrophobia.
or Afghanistan were recruited for this study, including 37
whose self-report of LOC and PTA associated with blast were
consistent with mild to moderate TBI. Seventeen service
members and veterans were recruited for the comparison
group, who reported no history of TBI and no exposure to
blast or LOC, including eight who reported at least one ex-
tracranial injury during deployment, and nine who were
uninjured (Table 1). Two subjects from the comparison group
were excluded due to incidental MRI findings (see below).
Imaging data from one of the TBI subjects were not analyzed
due to technical difficulties.
Magnetic resonance imaging.
quired on a 3.0 T Philips Achieva scanner (Cleveland, OH) at
the MEDVAMC using an 8-channel SENSE headcoil. Anato-
mical series to assess neuropathology included T1-weighted
3D-turbo field echo (9.9msec repetition time [TR], 4.6msec
echo time [TE], 1.0mm axial slices, 0mm gap, 132 slices,
240mm field of view [FOV], 88 flip angle), T2-weighted gra-
dient echo (2500msec TR, 32msec TE, 5.0mm axial slices,
0mm gap, 25 slices, 224mm FOV, 30–408 flip angle), T2-
weighted fluid attenuated inversion recovery (FLAIR)
(11000msec TR, 105msec TE, 5.0mm axial slices, 1mm gap,
25 slices, 240mm FOV, 908 flip angle), and T2-weighted gra-
dient spin echo (GRASE) imaging (2141msec TR, 80msec TE,
5.0mm axial slices, 0mm gap, 25 slices, 230mm FOV, 908 flip
Unsedated MRI was ac-
Diffusion tensor imaging.
echo, single shot, echo planar imaging (EPI) sequences were
used (7341msec TR, 70msec TE, 2.0mm axial slices, 0mm
gap). A 224mm FOV was used with a measured voxel size of
2.0?2.0?2.0mm, and a reconstructed voxel size of 1.75?
1.75?2.0mm. Diffusion was measured along 32 direc-
tions (number of b-value¼2, low b-value¼0, and high b-
value¼1000sec=mm2). To improve signal-to-noise ratio,
high-b images were acquired twice and averaged in most
cases. Each acquisition took approximately 4min 44sec, and
70 slices were acquired.
Transverse multi-slice spin
Review of magnetic resonance imaging findings.
images were reviewed by a board-certified neuroradiologist
at the MEDVAMC independently of group identification or
outcome data. The presence, location, and pathology of each
lesion were documented, and the presence and severity of
diffuse pathology such as atrophy were also coded. This re-
inferior frontal gyrus of one comparison participant, and
multiple areas of shearing injury within the frontal lobes of
another. These findings were interpreted as evidence of old
traumatic injury and both subjects were later excluded from
the overall analysis. Five veterans with TBI had MRI findings
consistent with brain trauma, and their neuropathology is
summarized in Table 2. One TBI participant had a Chiari type
I malformation as an incidental finding.
Diffusion tensor imaging analysis.
analysis was guided by the literature on the neuropathology
of TBI documenting vulnerability of parasagittal regions to
axonal injury (Adams et al., 1982; Meythaler et al., 2001), and
We measured FA and ADC for both manually-traced regions
of interest (ROIs) on a slice-by-slice basis, and quantitative
fiber tractography to assess WM integrity, as these methods
have different advantages and limitations.
Selection of sites for
Diffusion tensor imaging data preprocessing.
DTI registration tool (release 0.4; Netsch and van Muiswinkel,
2004) for Philips Research Integrated Development Environ-
ment V4 was used, with the following parameters: affine
transformation, 10% local correlation, maximum number of
steps¼20, minimum distance 50mm, removal of motion and
eddy current artifacts, and averaging two acquisitions for
better signal-to-noise ratio.
given any information concerning the group identification or
outcome data. The Philips fiber tracking software (Hoogen-
raad, 2002) was utilized to determine fiber tracts passing
through ROIs, and mean FA and ADC of the identified fibers
was used as the quantitative measure for DTI variables. The
methodology and quantitative tractography protocols for the
corpus callosum (including genu, body, splenium, and total),
arcuate fasciculus, inferior longitudinal fasciculus, inferior
fronto-occipital fasciculus, uncinate fasciculus, cingulate
bundle, and anterior and posterior limbs of the internal cap-
sule have been previously published (Levin et al., 2008; Wa-
kana et al., 2007; Wilde et al., 2006; Wilde et al., 2009). Figure 1
illustrates the quantitative tractography method used for the
corpus callosum and cingulate regions in a subject with blast-
related DTI. Corticospinal tracts were measured with a mul-
tiple ROI method in the axial plane, with one ROI placed at
the level of the colliculi, and the other placed at the WM just
superior to the roof of the lateral ventricles. The fornix was
tensor imagingquantitativefiber tracto-
Operators who analyzed the DTI data were not
Table 1. Demographic Features in the TBI and Comparison Groups
MeanSDSD Wilcoxon Z scoresp Value
Post-injury interval to scan (days)
TBI, traumatic brain injury; SD, standard deviation; IQ, intelligence quotient using the Barona et al. (1984) method of estimation.
DTI IN BLAST INJURY 685
measured using two seed point ROIs placed along the fornix
in two different coronal slices where the fornix body is visible
as one bundle.
Standard (non-tractography) diffusion tensor imaging
regions of interest.
Non-tractography ROI protocols in-
cluded the total corpus callosum (including genu, body,
splenium, and total), and the right and left anterior and pos-
terior limbs of the internal capsule. The total corpus callosum
and its subregions were measured on the mid-sagittal slice,
using the same boundaries as the quantitative tractography
measurement (Wilde et al., 2006). The boundaries used for the
internal capsule were also applied in a similar fashion to ob-
FA and ADC within the defined ROI on one slice was per-
formed. The location and boundaries of regions involved in
the ROI method are illustrated in Figure 1.
Intra-rater and inter-rater agreement.
rater agreement, analysis of each region was performed twice
for each participant for both quantitative tractography and
standard ROI measurement; intra-class correlation coeffi-
cients (ICCs) exceeded 0.95 for all DTI indices. Inter-rater
agreement was also assessed by measurement of each proto-
To examine intra-
Table 2. Durations of LOC=PTA and MRI Pathology in TBI Participants
of PTA Lesion location and pathology
24 hours–7 days
24 hours–7 days
Chiari type I malformation
(R=L) SFG and (R) FLWM
and MFG shearing injury
of WM and G=WJ
MDBA, (R=L) FLWM, SFG,
and PG shearing injury
of WM and G=WJ
(R=L) FLWM, SFG,
and MFG gliosis of WM
(R=L) SFG and IFG, (L)
PL encephalomalacia and
shearing injury of G=WMJ
7 days–1 month
7 days–1 month
MDBA, mild diffuse brain atrophy; FLWM, frontal lobe white matter; SFG, superior frontal gyrus; MFG, middle frontal gyrus; IFG, inferior
frontal gyrus; PG, precentral gyrus; PL, parietal lobe; WM, white matter; G=WMJ, grey=white matter junction; R, right; L, left; LOC, loss of
consciousness; PTA, post-traumatic amnesia; TBI traumatic brain injury; MRI, magnetic resonance imaging.
686LEVIN ET AL.
col by two different raters in five cases in each group; ICCs
again exceeded 0.95.
MD) to compute and export DTI index maps including ADC
and FA. To achieve better inter-subject registration, the point
of intersection of the anterior commissure with the mid-
sagittal plane was selected as the landmark and manually
reset as the origin of each exported map image. The FA and
ADC maps were calculated with the background noise
threshold of10,whichexcludedpixelswithlowerintensity on
the b¼0 image. In addition, b¼0 images were also exported.
The b¼0 image for one control participant was used to create
a template in Montreal Neurological Institute space. All DTI
then smoothed with a gaussian filter of full width at half
maximum of 8mm. After data preparation, the voxel-based
analysis (VBA) was performed using SPM2 (Wellcome De-
partment of Imaging Neuroscience, London, U.K.) running on
t-test was performed between TBI and comparison groups to
evaluate any difference in FA and ADC. In these analyses, a
cluster threshold of 10 and a p value of 0.01 were used.
The preprocessed DTI data were
Participants in both groups underwent outcomes assess-
ment at the MEDVAMC on the day of imaging or within
symptoms were evaluated in both groups by the Neurobe-
havioral Symptom Inventory (NSI) (Cicerone and Kalmar,
1995), a self-report symptom inventory of 22 post-concussion
symptoms. Inclusion of the non-TBI group provided an esti-
mate of the level of symptoms such as headache in a group of
post-deployment veterans. The participant rated the presence
and severity of each symptom for the past 2 weeks on a five-
point scaleof severity. Cluster analysis (Cicerone and Kalmar,
1995) of the NSI has revealed physical, cognitive, affective,
and somatic symptom clusters. The reliability and validity of
the NSI have been supported in previous studies (Cicerone
and Kalmar, 1995). We administered the PTSD Checklist-
Civilian Version (PCL-C) (Blanchard et al., 1996), a self-report
symptom inventory of 17 PTSD symptoms which the partic-
ipant rated on a five-point scale of severity based on the past
month. The reliability and validity of the PCL-C have been
found to be satisfactory (Blanchard et al., 1996). The total
PTSD severity score of the PCL-C was analyzed. From the
Brief Symptom Inventory (BSI) (Derogatis, 1982) we analyzed
the Global Severity Index and the depression scale. HRQOL
was measured by the Mental and Physical Summary scores of
SF-12, a brief version of the Medical Outcomes Study Short-
Form (SF-36) which is supported by reliability and validity
data (Ware et al., 1996).
using the Verbal Selective Reminding Test (VSRT; Buschke,
1973), and the Nonverbal Selective Reminding Test (NSVRT;
Fletcher, 1985). The verbal version tested consistent long-term
retrieval (CLTR) of 12 words over six trials, followed by a
delayed recall trial administered 30min later. The Nonverbal
Selective Reminding Test (Fletcher, 1985) measured CLTR of
the spatial location for eight dots over eight trials followed by
a 30-min delayed recall. Total number of items recalled con-
sistently across trials and the delayed recall score were ana-
lyzedfor both tests.To measuredecision makinginrelation to
rewards and penalties over trials, we administered the Iowa
Gambling Task (Bechara et al., 1994), in which the participant
decks. The participant was free to select a card from any of the
decks on each trial in response to rewards or penalties which
for this task was the somatic marker hypothesis that emo-
tional signals bias selection of advantageous responses and
inhibition of disadvantageous responses (Damasio, 1996).
Bechara and associates (2000) postulated that a neural system
consisting of ventral medial prefrontal cortex, amygdala, and
ventral striatum mediates the capacity to shift preference to
advantageous responses as experience accrues with rewards
and penalties. A rival postulation (Rolls, 1999) differs, mainly
by implicating the lateral orbital frontal cortex rather than the
ventral medial region. The decks differed in the schedule and
Episodic memory was evaluated
moderate blast-induced TBI (reportedly>30min duration of post-injury confusion), demonstrating both the quality of DTI
data acquisition and the protocol used for the anterior and posterior limbs of the internal capsule using the region-of-interest
(ROI) analysis method (boundaries are outlined in white). Next, the DTI FA color map for the same individual is portrayed in
the sagittal plane, and boundaries used for the subregions of the corpus callosum (genu, body, and splenium) using the ROI
analysis method are illustrated (boundaries are outlined in white). Third, a sagittal view of the tractography of the total
corpus callosum is overlaid on the same subject’s T1-weighted image, again illustrating the excellent quality of the tracto-
graphy results, as well as the characteristic fiber pattern that emerges using the quantitative tractography method of analysis.
Finally, we see a sagittal view of the tractography of the left cingulum bundle overlaid on the subject’s T1-weighted image.
Note that the tractography fiber patterns for this individual with a history of moderate blast-induced traumatic brain injury
do not qualitatively differ from the expected pattern observed in control subjects, consistent with the lack of quantitative
difference in the DTI metrics obtained.
From left to right: Axial diffusion tensor imaging (DTI) fractional anisotropy (FA) color map of a subject with
DTI IN BLAST INJURY 687
amount of rewards and penalties given for selecting a card.
a card from two of the decks (C and D) resulted in large
rewards on early trials, followed by costly penalties on later
trials (disadvantageous decks), whereas selection of cards
from the other two decks (A and B) produced consistent, al-
of cards from the advantageous decks generated larger net
gains over many trials. The difference between the sum of
cards selected from decks A and B and decks C and D was the
dependent measure. With a dichotomous variable of cards
selected from advantageous versus disadvantageous decks,
we also analyzed the distribution of this difference score and
Because of the small sample size and because some of the
normal approximation continuity-corrected z statistic for the
non-parametric Wilcoxon rank sum two-sample test to com-
pare the TBI and comparison groups on continuous variables.
The non-parametric Kruskal-Wallis test was used to compare
the mild TBI, moderate TBI, and comparison groups on con-
tinuous variables. A repeated-measures analysis was used to
and comparison groups on the Verbal Selective Reminding
and the Nonverbal Selective Reminding tests, which were
composed of six trials, and eight trials, respectively. For
between-group comparisons of categorical variables, we used
Fisher’s exact test for 2?2 tables, and chi-square test for
comparison of variables with more than two categories. Stu-
dent’s t-tests were used to test between-group differences in
the VBA of DTI variables. Correlation of imaging data with
the outcome measures was performed using the Spearman
rank-order statistic. p Values<0.01 were considered signifi-
Demographic and clinical features
The groups did not significantly differ in age, education,
pre-injury IQ estimated from demographic features (Barona
et al., 1984), or the time from the date of injury (or end of
deployment for uninjured veterans in the comparison group;
Asseen in Table 2,17 (46%) ofthe TBIparticipants reported
Eight TBI participants (22%) reported durations of PTA or
confusion longer than 24h, for a total of nine participants
(24%) whose duration of LOC, altered consciousness, or PTA
was consistent with moderate TBI. The remaining 28 TBI
participants reported durations of LOC or altered conscious-
ness indicative of mild TBI. For the 27 TBI participants who
reported exposure to more than one blast, PTA and LOC for
the most severe injury are represented in Table 2.
Post-concussion and symptoms of post-traumatic
Participants with TBI reported significantly more severe
physical, cognitive, and sensory clusters, but not for the af-
fective cluster, of the NSI (Table 3). The mean total PTSD
symptom score on the PCL-C was also significantly higher in
the TBI group (Table 4).
According to an established criterion T-score >50 on the
PCL-C, 22 of 37 (59.5%) of the TBI participants had scores
consistent with the diagnosis of PTSD, compared with 4 of 15
Total PTSD and the sum of the post-concussion symptom
scores (Table 3) were also highly correlated within the TBI
(Spearman r¼0.89, p<0.0001) and comparison (r¼0.94,
p<0.001) groups. When we adjusted the PTSD total score for
the sum of the post-concussion symptom cluster scores on the
NSI, the between-group difference was eliminated. Similarly,
adjusting the sum of the NSI cluster scores for the PTSD to-
tal score mitigated the between-group difference in post-
Emotional distress, depression, and health-related
quality of life
Global emotional distress and depression as measured by
the BSI were more severe in the TBI group than the compar-
ison group (Table 4). According to a criterion T score of 63 or
higher on the Global Severity Index of the BSI, 36 of the 37
(97.3%) TBI participants had generalized emotional distress,
compared with 9=15 (60.0%) of the comparison group
(Fisher’s exact probability¼0.0014). The groups did not sig-
nificantly differ in the Physical or Mental Summary scores of
the SF-12 (Table 4).
Between-group analysis of total CLTR across trials dis-
closed that verbal learning and memory on the Verbal Selec-
tive Reminding Test was less efficient in the TBI group than
Table 3. Mean Post-Concussion Cluster Scores for Neurobehavioral Symptom Inventory
MeanSDMean RangeRange Wilcoxon statisticp Value
Sum of mean cluster scores
TBI, traumatic brain injury; SD, standard deviation.
688LEVIN ET AL.
the comparison participants (F (1,51)¼22.92, p<0.0001). As
seen in Figure 2, delayed recall of the word list was also lower
in the TBI group than the comparison group (F (1,51)¼14.18,
p¼0.0004). In contrast, nonverbal learning and memory
(CLTR) for the spatial location of dots across trials (F (1,43)¼
2.23, p¼0.143), and recall after a delay (F (1,43)¼0.91,
p¼0.345) did not differ between the groups.
Spearman correlations between the verbal memory vari-
ablesandthetotalPTSD score werenon-significant,including
the total number of words consistently retrieved across trials
(r¼–0.111, p¼0.52), and delayed recall (r¼?0.202, p¼0.23).
For Nonverbal Selective Reminding, correlations with total
PTSD score were r¼?0.293 (p¼0.10) for consistent retrieval,
and r¼?0.317 (p¼0.08) for delayed recall of spatial loca-
Decision-making performance on the IowaGambling Task,
as measured by the difference in the number of cards selected
from advantageous and disadvantageous decks over 100 tri-
als, did not differ between the TBI (mean¼3.65, SD¼23.74)
and the comparison (mean¼8.0, SD¼26.33) groups (F
(1,47)¼0.33, p¼0.570), and no between-group differences in
Diffusion tensor imaging findings
Distributions of FA and ADC as measured by fiber tracto-
graphy or the ROI method were similar for the TBI and
comparison groups. For example, fiber tractography dis-
closed a mean FA for the total corpus callosum of 0.50
(SD¼0.02, range¼0.46–0.54) for the TBI group, and 0.50
(SD¼0.02, range¼0.45–0.53) for the comparison group,
Wilcoxon statistic¼?1.06, p¼0.48. Corresponding ADC
values for the TBI group (mean¼0.84, SD¼0.03, range¼
0.79–0.89) did not differ from those of the comparison group
(mean¼0.85, SD¼0.06, range¼0.77–0.97), Wilcoxon statis-
Consistent with this pattern, no between-group differences
were found on either ADC or FA using either fiber tracto-
graphy or non-tractography methods, including a single-slice
ROI approach and a VBA.
To further examine the possibility that the lack of group
differences on DTI parameters may have been attributable to
the relatively mild nature of the blast-injury exposure, we
repeated analysis of DTI variables for both quantitative trac-
tography and ROI methods by comparing the data of the 27
subjects with mild TBI with findings for nine subjects with
moderate TBI in whom severity of injury was defined by
duration of LOC, altered consciousness, or PTA (Table 2),
relative to the 15 subjects in the comparison group using
Kruskal-Wallis tests. For quantitative tractography, there
were no significant differences found among these groups on
any DTI variable. Using the ROI method, the FA of the total
corpus callosum marginally differed among the groups (w2
(2)¼6.16, p¼0.05), but contrary to expectations, the moder-
ate TBI group demonstrated a higher mean FA in this region
than either the mild TBI or the comparison group.
Correlation of diffusion tensor imaging
findings with outcome measures
We found no significant correlations between the outcome
measures and DTI variables within the comparison group of
veterans and service members who did not sustain a TBI and
were not exposed to blast. As described below, the direction,
magnitude, and consistency of these brain–behavior correla-
tions within the TBI group varied with the outcome domain
Table 4. Mean Scores for GSI and Depression on the BSI, PTSD on PCL-C, and SF-12
SDMean RangeMeanRange Wilcoxon statisticb
Depression Scale (BSI)a
PTSD score (PCL-C)
aThe BSI variables are T-scores for non-patient norms.
bThe value for the Wilcoxon statistic is the normal approximation continuity-corrected z statistic.
TBI, traumatic brain injury; SD, standard deviation; BSI, Brief Symptom Inventory; GSI, Global Severity Index; PTSD, post-traumatic stress
disorder; PCL-C, Posttraumatic Stress Disorder Checklist-Civilian Version; SF-12, Medical Outcomes Study Short Form (SF-12) Physical and
Mental Summary Scores.
trials on the Verbal Selective Reminding Test, and following
a 30-min delay, for both the traumatic brain injury (TBI) and
comparison groups (CLTR, consistent long-term retrieval).
Number of words consistently recalled over six
DTI IN BLAST INJURY689
between the DTI variables of FA and ADC and the total post-
concussion symptom score on the NSI and the total PTSD
symptom score on the PCL-C were calculated for the TBI
group. Using fiber tracking, total corpus callosum FA was
negatively correlated with both the mean post-concussion
symptom cluster score (r¼?0.40, p¼0.02), and the mean
total PTSD score (r¼?0.30, p¼0.07). However, there were no
significant correlations of ADC with the symptom measures,
and no significant correlations of either FA or ADC using the
Spearman rank-order correlations
Table 5, the most consistent correlations with verbal memory
on the selective reminding tests were obtained for the unci-
nate, internal capsule, and corticospinal tract. FA values in-
dicating integrity of WM microstructure were positively
related to consistent verbal recall across trials, whereas ADC
values reflecting increased diffusivity were negatively related
to recall (Fig. 3). However, the corresponding correlations for
As seen in the fiber tracking data in
as both FA and ADC values were positively correlated with
consistency of recall and delayed recall, respectively (Table 5).
For the ROI method in the TBI group, Table 6 shows pos-
itive correlations between FA in the splenium of the corpus
callosum and the number of words consistently retrieved
across trials and after a delay, consistent with the expected
pattern in which higher FA is related to better performance
(Fig. 3). Negative correlations were found between the total
number of words consistently retrieved and ADC in the left
posterior limb and the right anterior limb of the internal
capsule. Again this was consistent with the expected pattern
in which higher ADC was related to poorer verbal memory.
No significant correlations were found between DTI variables
obtained with the ROI method and either consistent recall
across trials or delayed recall on the NVSRT.
formance on the Iowa Gambling Task and DTI variables
were limited to the quantitative fiber tracking method, in-
cluding right uncinate FA (r¼0.426, p¼0.003), right posterior
internal capsule FA (r¼0.325, p¼0.024), and right inferior
Significant correlations between per-
Table 5. Correlations of FA and ADC Measured by Fiber Tracking and Verbal and Nonverbal
Selective Reminding Tests
Total CLTR VSRT Total CLTR NVSRT 30min delayed recall NVSRT
Right PIC FA
Left PIC FA
Left CST FA
Right uncinate ADC
Left uncinate ADC
Left PIC ADC
Left AIC ADC
CC genu ADC
CC splenium ADC
CC total ADC
Right IFOF ADC
FA, fractional anisotropy; ADC, apparent diffusion coefficient; AIC, anterior limb of the internal capsule; CC, corpus callosum; CST,
corticospinal tract; IFOF, inferior fronto-occipital fasciculus; PIC, posterior limb of the internal capsule; CLTR, consistent long-term retrieval;
NVSRT, Nonverbal Selective Reminding Test; VSRT, Verbal Selective Reminding Test.
(FA) and apparent diffusion coefficient (ADC) as measured by diffusion tensor imaging using fiber tracking (FT) and region-
of-interest (ROI) methods. Correlation coefficients are shown in Tables 5 and 6 (CLTR, consistent long-term retrieval).
Scatterplots of consistent retrieval of words on the Verbal Selective Reminding Test plotted by fractional anisotropy
690LEVIN ET AL.
frontal occipital fasciculus FA (r¼0.325, p¼0.024). There were
no significant correlations with ADC or using ROI measures.
Post-concussion and symptoms of post-traumatic
Our outcomedatareplicate those of previous studies (Hoge
et al., 2008; Schneiderman et al., 2008), showing that OEF=OIF
veterans and service members with a history of blast-induced
mild to moderate TBI report more severe post-concussion and
PTSDsymptoms thanveteranswithout TBI.Basedon theNSI,
than the comparison group, a finding that was confirmed for
physical, cognitive, and sensory clusters of post-concussion
the affective symptom cluster of the NSI. Nearly 60% of par-
ticipants with TBI had elevated total PCL-C scores consistent
with the diagnosis of PTSD, compared with a previously re-
ported rate of 50% using the same diagnostic criterion among
OEF=OIF returnees who sustained mild blast-related TBI with
previous studies, we found that OEF=OIF veterans with mild
TBI had PTSD symptoms that were more severe than veterans
who had been deployed but had no TBI or exposure to blast.
The linear relation between severity of post-concussion and
PTSDsymptoms in ourTBIparticipants implicatessubstantial
overlap and raises the possibility of a common, unifying
mechanism. This overlap of symptoms is consistent with self-
report data collected from a large cohort of veterans who had
sustained blast-induced TBI, predominantly in the mild range
(2008), we found that increased post-concussion symptoms in
by adjusting for severity of PTSD symptom scores. Con-
versely, adjusting for post-concussion symptoms eliminated
the between-group difference in PTSD symptom scores. Our
findings are consistent with the interpretation by Hoge and
colleagues (2008), that blast-induced mild TBI and PTSD are
frequently comorbid in OEF=OIF veterans and service mem-
bers. In comparison with civilian mild TBI, blast-related mild
TBI in the OEF=OIF population is often associated with an-
ticipatory emotional arousal and exposure to deaths and in-
juries among other service members, enemy combatants, or
noncombatants, thus increasing the risk for PTSD (Vander-
ploeg et al., 2009; Vasterling et al., 2006). Moreover, this
traumatic exposure may be repetitive, further increasing the
risk for PTSD in OEF=OIF veterans with mild TBI.
Global distress and health-related quality of life
Global emotional distress as measured by the BSI was also
more severe in the TBI group relative to the comparison
group. However, the groups differed only marginally on the
depression scale score of the BSI, as clinically elevated
depression scores were obtained in more than half of the
participants in the comparison group, thus mitigating a sig-
nificant between-group difference. This marginal finding is
consistent with other reports (Invisible Wounds of War, 2008;
Seal et al., 2009) of generally high rates of depression in
OEF=OIF veterans. In contrast, we found no between-group
difference in HRQOL as measured by self-perceived physical
health and mental health concerns on the SF-12. The dissoci-
ation in findings between measures of symptoms, distress,
and HRQOL, suggests that the increased symptoms associ-
ated with TBI were disproportionately severe for the post-
concussion and post-traumatic stress domains.
We found that group differences in cognition were specific
to the VSRT (Buschke, 1973), a measure of episodic memory
that involves learning a list of 12 words and resistance to
interference. Followingthefirst trial,inwhich all12wordsare
presented, the examiner presented only those words that the
participantfailedtorecall ontheprecedingtrial. Thisselective
reminding thus potentially interferes with recall of those
words that were correctly recalled on the previous trial and
thus not represented by the examiner. Veterans and service
members with blast-related TBI recalled fewer words across
the six trials and following a 30-min delay than the compar-
ison group. However, we found no between-group differ-
ences in visual recall of the spatial location of dots using a
selective reminding procedure (Fletcher, 1985). This dissoci-
ation is in accord with the sensitivity of the VSRT to civilian
mild TBI (Levin et al., 1987). The finding of poor verbal
memory more than a year after blast-related mild TBI con-
trasts with the recovery typically found on this test within 1–3
months after civilian mild TBI (Levin et al., 1987). With the
exception of sports-related mild TBI (Guskiewicz et al., 2003),
civilians typically sustain a single injury, whereas repetitive
exposure to blast is common in OEF=OIF service members,
especially those undergoing multiple deployments (Vasterl-
ing et al., 2006). The potentially cumulative effects of repeti-
tive exposure to blast further differentiate the injuries
sustained by OEF=OIF veterans from civilian mild TBI. The
association of PTSD with problems in initial registration of
information, vulnerability to interference, and sustained
had poor verbal memory using a selective reminding proce-
dure. This postulation is also congruent with the report
(Brewin et al., 2010) that verbal memory is more vulnerable
than visual memory to PTSD (Vanderploeg et al., 2009), and
the view (Vanderploeg et al., 2009) that mild TBI due to blast
and PTSD independently contribute to cognitive deficits.
Administration of the Iowa Gambling Task (Bechara et al.,
1994) to evaluate decision making in relation to incentive and
penalties disclosed no between-group difference in the rela-
tive preference for selecting cards from advantageous decks
across trials. Moreover, the distributions of the difference
between total cards selected from advantageous versus dis-
advantageous decks did not differ, and outliers were no more
Table 6. Correlations of FA and ADC as Measured
by ROI and the Verbal Selective Reminding Test
Total CLTR30-Min delayed recall
CC splenium FA
Left PIC ADC
Right AIC ADC
FA, fractional anisotropy; ADC, apparent diffusion coefficient;
AIC, anterior limb of internal capsule; CC, corpus callosum; PIC,
posterior limb of internal capsule; CLTR, consistent long-term
retrieval; ROI, region of interest.
DTI IN BLAST INJURY691
frequent in the mild TBI group than the comparison group.
Without evidence for a deficiency in deferring reward to
maximize gains over trials, our data provide no support for
dysfunction in the neural system involving the ventral pre-
frontal cortex, amygdala, ventral striatum, and nucleus ac-
cumbens in OEF=OIF veterans with mild TBI due to blast.
Diffusion tensor imaging findings
Using manual measurement of single-slice ROIs, quanti-
tative tractography, and VBA methods of DTI analysis, we
found no between-group differences in the integrity of WM
microstructure. Moreover, the distributions of FA and ADC
were quite similar in the two groups across neuroanatomic
regions of WM known to be vulnerable to axonal injury, and
correlations between the DTI data and post-concussion
symptoms were generally weak. The lack of between-group
differences in DTI indices may be attributed to the WM injury
having been of subthreshold severity for detection by this
imaging technique. With the exception of nine participants
who reported duration of LOC longer than 30min, or dura-
tions of PTA or confusion longer than 24h, our sample of TBI
participants had LOC and PTA consistent with mild TBI.
However, we found no consistent difference in DTI variables
between the mild and moderate TBI subgroups. Moreover,
disclosed no consistentpattern,
concussion and PTSD symptoms. We also evaluated whether
high FA, reflecting integrity of WM microstructure, was
positively correlated with cognitive performance within the
TBI group. As a convergent analysis, we determined whether
increased diffusivity as reflected by higher ADC values was
negatively correlated with cognitive performance. Although
the direction of correlations for the VSRT was consistent with
expectations, and reflected increased memory problems with
lower FA and higher ADC, this was not the case for the
NVSRT, for which the direction of the correlations was in-
consistent and at times in an unanticipated direction. We also
found no significant correlation between DTI variables and
severity of PTSD symptoms, indicating that post-traumatic
stress could not explain the relation of verbal memory to DTI
indices. Despite the lack of a correlation between decision
making on the Iowa Gambling Task and ADC, performance
of the TBI group was positively related to FA in tracts con-
necting the prefrontal and temporal regions, and extending
from the prefrontal to occipital regions. Consistent with evi-
dence of greater sensitivity of right than left ventral medial
prefrontal lesions to poor decision making on this task (Be-
chara, 2004), the positive correlations with FA were found for
the right uncinate, right inferior frontal occipital fasciculus,
and the posterior limb of the right internal capsule in the TBI
group. Consequently, we suggest that further investigation of
this cognitive measure in relation to DTI may be informative.
Late DTI findings of reduced FA and increased ADC in-
dicative of WM injury in civilian TBI have generally been
confined to participants with moderate to severe injury. Al-
though a recent study (Niogi et al., 2008) has disclosed subtle
alteration ofthe relation between DTI indices and cognition in
chronic civilian mild TBI participants, between-group differ-
injury severity interpretation of our DTI data could be tested
in future DTI studies by including veterans with clinician-
including both post-
confirmed moderate to severe blast-induced TBI, in addition
to those with mild TBI. Longitudinal imaging, including pre-
deployment DTI, and repeated DTI within a week after mild
TBI (Bazarian et al., 2007; Wilde et al., 2008) due to blast
would also contribute to identifying subtle WM alterations
(Bazarian et al., 2007; Wilde et al., 2008) that may resolve over
time. Comparison with prospective civilian TBI patients
would also elucidate whether the negative DTI findings in the
present study are specific to mild blast-induced TBI. Neuro-
pathology studies support a continuum of WM injury in
civilian TBI (Blumbergs et al., 1994; Oppenheimer, 1968), but
the pattern seen in blast-related TBI remains to be elucidated.
Limitations of this study
In our sample of veterans and service members with
chronic blast-related TBI, assessment of acute injury variables
depended in part on self-report data concerning alteration or
loss of consciousness and PTA. However, the diagnosis of TBI
was based on a clinical interview with a physician, similar to
the approach used in other studies of veterans with TBIdue to
blast (Belanger et al., 2009; Terrio et al., 2009). We acknowl-
edge the small sample size of OEF=OIF veterans and service
members whom we recruited prospectively from a trauma
registry, from clinical services at the MEDVAMC, and from
advertising in the newspaper of a regional Army post. Al-
thoughrecruitment involved cooperationwithseveralclinical
services, study brochures were widely distributed and en-
rollment was not based on patients referred by clinicians be-
cause of more severe or persistent symptoms (Terrio et al.,
2009). Despite the small sample size, the behavioral findings
were consistent with those of previous studies (Hoge et al.,
2008; Schneiderman et al., 2008) using similar self-report
measures of post-concussion and PTSD symptoms. Also,
medical records documenting acute injury were not available.
Our comparison group of post-deployment OEF=OIF veter-
ans unexposed to blast was divided between those who sus-
tained extracranial injury and those who were uninjured.
Inclusion of uninjured veterans in the comparison group may
have accentuated the group differences in reported symp-
toms. The cross-sectional design of this study is a further
limitation. Recruitment through the trauma registry, clinics,
and Army base provided access to veterans and service
members with chronic, but not acute TBI. Although our
validity, group differences in cognitive performance were
specific to verbal recall. A dichotomous variable measuring
decision making over trials in relation to rewards and pen-
alties showed similar distributions of data for the TBI and
comparison groups. We also found no between-group dif-
ference in self-perceived health status, a finding inconsistent
with symptom amplification or malingering in the TBI group.
statistical tests, we view the correlation analyses between DTI
variables and outcome measures as exploratory in this initial
DTI study of blast-related TBI.
Further investigation of blast-induced TBI using DTI is
nevertheless encouraged, because this noninvasive imaging
technique is generally sensitive to civilian TBI injury due to
closed head trauma, and it provides a measure of WM
692LEVIN ET AL.
integrity independent of self-report or behavioral data. Sup-
and other advanced MR techniques such as magnetization
et al., 1995), and susceptibility-weighted imaging (Tong et al.,
mild blast-induced TBI and the concomitant effects of PTSD.
Moreover, longitudinal investigation with repeat imaging
would be more sensitive to the subtle changes that occur
following these injuries.
This research was supported by grant B4596 by the De-
partment of Veterans Affairs, Veterans Health Administra-
tion, Rehabilitation Research and Development Service; the
Mental Illness Research, Education and Clinical Centers
(MIRECC); and by the Houston VA HSR&D Center of Ex-
cellence (HFP90-020). We gratefully acknowledge the assis-
tance of Stacey K. Martin and Stephen R. McCauley, Ph.D. for
their assistance in manuscript preparation. We are also in-
debted to the veterans and service members who participated
in the study.
Author Disclosure Statement
No competing financial interests exist.
Adams, J.H., Graham, D.I., Murray, L.S., and Scott, G. (1982).
Diffuse axonal injury due to nonmissile head injury in hu-
mans: an analysis of 45 cases. Ann. Neurol. 12, 557–563.
Bagley, L.J., McGowan, J.C., Grossman, R.I., Sinson, G., Ko-
tapka, M., Lexa, F.J., Berlin, J.A., and McIntosh, T.K. (2000).
Magnetization transfer imaging of traumatic brain injury. J.
Magn. Reson. Imaging 11, 1–8.
Barona, A., Reynolds, C.R., and Chastain, R. (1984). A demo-
graphically based index of premorbid intelligence for the
WAIS-R. J Consult. Clin. Psychol. 5, 885–887.
Bazarian, J.J., Zhong, J., Blyth, B., Zhu, T., Kavcic, V., and Pe-
terson, D. (2007). Diffusion tensor imaging detects clinically
important axonal damage after mild traumatic brain injury: a
pilot study. J. Neurotrauma 24, 1447–1459.
Bechara, A. (2004). The role of emotion in decision-making: ev-
idence from neurological patients with orbitofrontal damage.
Brain Cogn. 55, 30–40.
Bechara, A., Damasio, A.R., Damasio, H., and Anderson, S.W.
(1994). Insensitivity to future consequences following damage
to human prefrontal cortex. Cognition 50, 7–15.
Bechara, A., Tranel, D., and Damasio, H. (2000). Characteriza-
tion of the decision-making deficit of patients with ventro-
medial prefrontal cortex lesions. Brain 123, 2189–2202.
Belanger, H.G., Kretzmer, T., Yoash-Gantz, R., Pickett, T., and Tu-
pler, L.A. (2009). Cognitive sequelae of blast-related versus other
mechanisms of brain trauma. J. Int. Neuropsychol. Soc. 15, 1–8.
Benson, R.R., Meda, S.A., Vasudevan, S., Kou, Z., Govindarajan,
K.A., Hanks, R.A., Millis, S.R., Makki, M., Latif, Z., Coplin, W.,
Meythaler, J., and Haacke, E.M. (2007). Global white matter
analysis of diffusion tensor images is predictive of injury se-
verity in traumatic brain injury. J. Neurotrauma 24, 446–459.
Blanchard, E.B., Jones-Alexander, J., Buckley, T.C., and Forneris,
C.A. (1996). Psychometric properties of the PTSD Checklist
(PCL). Behav. Res. Ther. 34, 669–673.
Blumbergs, P.C., Scott, G., Manavis, J., Wainwright, H., Simp-
son, D.A., and McLean, A.J. (1994). Staining of amyloid pre-
cursor protein to study axonal damage in mild head injury.
Lancet 344, 1055–1056.
Brewin, C.R., Gregory, J.D., Lipton, M.L., and Burgess, N. (2010).
Intrusive images in psychological disorders: Characteristics,
neural mechanisms, and treatment implications. Psychol. Rev.
Buschke, H. (1973). Selective reminding for analysis of memory
and learning. J. Verbal Learning Verbal Behav. 12, 543–550.
Cernak, I., and Noble-Haeusslein, L.J. (2009). Traumatic brain
injury: an overview of pathobiology with emphasis on mili-
tary populations. J. Cereb. Blood Flow Metab. 30, 255–266.
Champion, H.R., Holcomb, J.B., and Young, L.A. (2009). Injuries
from explosions: physics, biophysics, pathology, and required
research focus. J. Trauma 66, 1468–1477; discussion 1477.
Chu, Z., Wilde, E.A., Hunter, J.V., McCauley, S.R., Bigler, E.D.,
Troyanskaya, M., Yallampalli, R., Chia, J.M., and Levin, H.S.
(2010). Voxel-based analysis of diffusion tensor imaging in
mild traumatic brain injury in adolescents. AJNR Am. J.
Neuroradiol. 31, 340–346.
Cicerone, K.D., and Kalmar, K. (1995). Persistent postconcussion
syndrome: The structure of subjective complaints after mild
traumatic brain injury. J. Head Trauma Rehabil. 10, 1–17.
Damasio, A.R. (1996). The somatic marker hypothesis and the
possible functions of the prefrontal cortex. Philos. Trans. R.
Soc. Lond. B Biol. Sci. 351, 1413–1420.
Derogatis, L.R. 1982. The Brief Symptom Inventory: Adminis-
tration, Scoring, and Procedures Manual. Johns Hopkins
University School of Medicine: Baltimore.
Fletcher, J.M. (1985). Memory for verbal and nonverbal stimuli
in learning disability subgroups: analysis by selective re-
minding. J. Exp. Child Psychol. 40, 244–259.
French, L.M., and Parkinson, G.W. (2008). Assessing and treat-
ing veterans with traumatic brain injury. J. Clin. Psychol. 64,
Galarneau, M.R., Woodruff, S.I., Dye, J.L., Mohrle, C.R., and
Wade, A.L. (2008). Traumatic brain injury during Operation
Iraqi Freedom: findings from the United States Navy-Marine
Corps Combat Trauma Registry. J. Neurosurg. 108, 950–957.
Guskiewicz, K.M., McCrea, M., Marshall, S.W., Cantu, R.C.,
Randolph, C., Barr, W., Onate, J.A., and Kelly, J.P. (2003).
Cumulative effects associated with recurrent concussion in
collegiate football players: the NCAA Concussion Study.
JAMA 290, 2549–2555.
Hoge, C.W., McGurk, D., Thomas, J.L., Cox, A.L., Engel, C.C.,
and Castro, C.A. (2008). Mild traumatic brain injury in U.S.
Soldiers returning from Iraq. N. Engl. J. Med. 358, 453–463.
Hoogenraad, F. (2002). Multi-center evaluation of in-vivo fiber-
tracking. Philips Medical Systems.
Inglese, M., Makani, S., Johnson, G., Cohen, B.A., Silver, J.A.,
Gonen, O., and Grossman, R.I. (2005). Diffuse axonal injury in
mild traumatic brain injury: a diffusion tensor imaging study.
J. Neurosurg. 103, 298–303.
Invisible Wounds of War: Psychological and Cognitive Injuries,
Their Consequences, and Services to Assist Recovery. (2008).
T. Tanielian, and L.H. Jaycox, (eds), RAND Corporation: Santa
Kimura, H., Meaney, D.F., McGowan, J.C., Grossman, R.I., Len-
kinski, R.E., Ross, D.T., McIntosh, T.K., Gennarelli, T.A., and
Smith, D.H. (1996). Magnetization transfer imaging of diffuse
axonal injury following experimental brain injury in the pig:
characterization by magnetization transfer ratio with histo-
pathologic correlation. J. Comput. Assist. Tomogr. 20, 540–546.
DTI IN BLAST INJURY 693
Kraus, M.F., Susmaras, T., Caughlin, B.P., Walker, C.J., Sweeney, Download full-text
J.A., and Little, D.M. (2007). White matter integrity and cog-
nition in chronic traumatic brain injury: a diffusion tensor
imaging study. Brain. 130, 2508–2519.
Levin, H.S., Mattis, S., Ruff, R.M., Eisenberg, H.M., Marshall,
L.F., Tabaddor, K., High, W.M., Jr., and Frankowski, R.F.
(1987). Neurobehavioral outcome following minor head inju-
ry: a three-center study. J. Neurosurg. 66, 234–243.
Levin, H.S., Wilde, E.A., Chu, Z., Yallampalli, R., Hanten, G.R.,
Li, X., Chia, J., Vasquez, A.C., and Hunter, J.V. (2008). Diffu-
sion tensor imaging in relation to cognitive and functional
outcome of traumatic brain injury in children. J. Head Trauma
Rehabil. 23, 197–208.
Maisto, S.A., Carey, M.P., Carey, K.B., Gordon, C.M., and
Gleason, J.R. (2000). Use of the AUDIT and the DAST-10 to
identify alcohol and drug use disorders among adults with a
severe and persistent mental illness. Psychol. Assess. 12, 186–
Meythaler, J.M., Peduzzi, J.D., Eleftheriou, E., and Novack, T.A.
(2001). Current concepts: diffuse axonal injury-associated
traumatic brain injury. Arch. Phys. Med. Rehabil. 82, 1461–
Netsch, T., and van Muiswinkel, A. (2004). Quantitative evalu-
ation of image-based distortion correction in diffusion tensor
imaging. IEEE Trans. Med. Imaging 23, 789–798.
Niogi, S.N., Mukherjee, P., Ghajar, J., Johnson, C.E., Kolster, R.,
Lee, H., Suh, M., Zimmerman, R.D., Manley, G.T., and
McCandliss, B.D. (2008). Structural dissociation of attentional
control and memory in adults with and without mild trau-
matic brain injury. Brain 131, 3209–3221.
Oppenheimer, D.R. (1968). Microscopic lesions in the brain fol-
lowing head injury. J. Neurol. Neurosurg. Psychiatry 31, 299–
Owens, B.D., Kragh, J.F., Jr., Wenke, J.C., Macaitis, J., Wade,
C.E., and Holcomb, J.B. (2008). Combat wounds in Operation
Iraqi Freedom and Operation Enduring Freedom. J. Trauma
Povlishock, J.T., and Katz, D.I. (2005). Update of neuropathology
and neurological recovery after traumatic brain injury. J. Head
Trauma Rehabil. 20, 76–94.
Rolls, E.T. 1999. The Brain and Emotion. Oxford University Press:
Saunders, J.B., Aasland, O.G., Babor, T.F., de la Fuente, J.R., and
Grant, M. (1993). Development of the Alcohol Use Disorders
Identification Test (AUDIT): WHO Collaborative Project on
Early Detection of Persons with Harmful Alcohol Consump-
tion—II. Addiction 88, 791–804.
Schneiderman, A.I., Braver, E.R., and Kang, H.K. (2008). Un-
derstanding sequelae of injury mechanisms and mild trau-
matic brain injury incurred during the conflicts in Iraq and
Afghanistan: persistent postconcussive symptoms and post-
traumatic stress disorder. Am. J. Epidemiol. 167, 1446–1452.
Seal, K.H., Metzler, T.J., Gima, K.S., Bertenthal, D., Maguen, S.,
and Marmar, C.R. (2009). Trends and risk factors for mental
health diagnoses among Iraq and Afghanistan veterans using
Department of Veterans Affairs health care, 2002–2008. Am. J.
Public Health. 99, 1651–1658.
Skinner, H.A. (1982). The drug abuse screening test. Addict
Behav. 7, 363–371.
Smith, D.H., Meaney, D.F., Lenkinski, R.E., Alsop, D.C., Gross-
man, R., Kimura, H., McIntosh, T.K., and Gennarelli, T.A.
(1995). New magnetic resonance imaging techniques for the
evaluation of traumatic brain injury. J. Neurotrauma 12, 573–
K., Scally, K., Bretthauer, R., and Warden, D. (2009). Traumatic
brain injury screening: preliminary findings in a US Army Bri-
gade Combat Team. J. Head Trauma Rehabil. 24, 14–23.
Tong, K., Holshouser, B., Shutter, L., Chiou, P., Herigauk, G.,
and Haacke, E. (2002). High resolution susceptibility weighted
imaging (SWI) improves detection of hemorrhagic lesions in
adults with traumatic brain injury: correlation with severity of
injury and outcome, in: Proceedings of ISMRM 10th Scientific
Meeting, p. 46.
Vanderploeg, R.D., Belanger, H.G., and Curtiss, G. (2009). Mild
traumatic brain injury and posttraumatic stress disorder and
their associations with health symptoms. Arch. Phys. Med.
Rehabil. 90, 1084–1093.
Vasterling, J.J., Proctor, S.P., Amoroso, P., Kane, R., Heeren, T.,
and White, R.F. (2006). Neuropsychological outcomes of army
personnel following deployment to the Iraq war. JAMA 296,
Wakana, S., Caprihan, A., Panzenboeck, M.M., Fallon, J.H.,
Perry, M., Gollub, R.L., Hua, K., Zhang, J., Jiang, H., Dubey,
P., Blitz, A., van Zijl, P., and Mori, S. (2007). Reproducibility of
quantitative tractography methods applied to cerebral white
matter. Neuroimage 36, 630–644.
Warden, D. (2006). Military TBI during the Iraq and Afghanistan
wars. J. Head Trauma Rehabil. 21, 398–402.
Ware, J., Jr., Kosinski, M., and Keller, S.D. (1996). A 12-Item
Short-Form Health Survey: construction of scales and pre-
liminary tests of reliability and validity. Med. Care. 34, 220–
Wilde, E.A., Chu, Z., Bigler, E.D., Hunter, J.V., Fearing, M.A.,
Hanten, G., Newsome, M.R., Scheibel, R.S., Li, X., and Levin,
H.S. (2006). Diffusion tensor imaging in the corpus callosum in
children after moderate to severe traumatic brain injury. J.
Neurotrauma 23, 1412–1426.
Wilde, E.A., McCauley, S.R., Chu, Z., Hunter, J.V., Bigler, E.D.,
Yallampalli, R., Wang, Z.J., Hanten, G., Li, X., Ramos, M.A.,
Sabir, S.H., Vasquez, A.C., Menefee, D., and Levin, H.S.
(2009). Diffusion tensor imaging of hemispheric asymmetries
in the developing brain. J. Clin. Exper. Neuropsychol. 31, 205–
Wilde, E.A., McCauley, S.R., Hunter, J.V., Bigler, E.D., Chu, Z.,
Wang, Z.J., Hanten, G.R., Troyanskaya, M., Yallampalli, R., Li,
X., Chia, J., and Levin, H.S. (2008). Diffusion tensor imaging of
acute mild traumatic brain injury in adolescents. Neurology
Wu, T., Wilde, E.A., Bigler, E.D., Yallampalli, R., McCauley,
S.R., Troyanskaya, M., Chu, Z., Li, X., Hanten, G., Hunter, J.V.,
and Levin, H.S. (2009). Evaluating the relation between
memory functioning and cingulum bundles in acute mild
traumatic brain injury using diffusion tensor imaging. J.
Neurotrauma 27, 303–307.
Address correspondence to:
Harvey S. Levin, Ph.D.
Baylor College of Medicine
1709 Dryden Road, Suite 1200
Houston, TX 77030
694 LEVIN ET AL.