Table 3.

DTI Studies in Mild TBI

First Author	Year	Type of Study (time post-injury)	Magnet	Subjects (N, gender, age)	DTI Analysis Method and Dependent Measures	Brain Region(s)	Main Finding(s)
Arfanakis	2002	Acute. (24hrs). In 2 patients, baseline compared with DTI at 1 month follow-up.	1.5T	Patients: 5 patients (3M, 2F; mean age 35.6) [GCS 13–15]. Conventional CT normal. Controls: 10 controls (5M, 5F; mean age 28.9 years).	Analysis and Methods: ROI Analysis. Dependent Measures: Trace, FA and LI. (Lattice Index).	5 ROI structures with selected volumes (i.e., representative only), in the left and right hemispheres: Anterior and posterior CC, external capsule, anterior and posterior internal capsule.	Patients but not controls showed FA differences for ROIs between homologous hemispheres. Reduced FA in patients in the internal capsule and CC compared to controls. No trace differences reported. At follow-up two patients showed improvement in some but not all areas.
Inglese	2005	Acute and Chronic. In 20 of the patients, imaging was performed at a mean of 4.05 days post- injury. In the remaining 26, imaging was performed a mean of 5.7 years post- injury.	1.5T	Patients: 46 patients (29M, 17F; mean age 36) [GCS 13–15]. Conventional MRI showed contusions in 5 patients and hematomas in 3. Controls: 29 age- and sex-matched controls (15 M, 14 F; mean age 35). Negative findings on conventional MRI.	Analysis and Methods: Whole-Brain Histogram Analysis and ROI Analysis. Dependent Measures: FA and MD.	3 ROI structures with selected voxels (i.e., representative, only): the centrum semiovale, CC, and internal capsule.	No differences between groups using histogram-derived measures. Patients showed reduced FA in the CC, internal capsule, and centrum semiovale, and increased MD in the CC and internal capsule, compared to controls. MD in splenium of CC was higher in patients scanned earlier but lower in the posterior limb of the internal capsule, compared to patients scanned later.
Bazarian	2007	Acute. 72 hours: post- concussive (PCS) and neurobehavioral testing. At 1 month: quality of life and PCS assessments (n=11).	3T	Patients: 6 subjects [GCS 13–15]. Conventional CT normal. Controls: 6 age- and sex- matched orthopedic controls; Conventional CT normal. Subjects (patients and controls) included 8M and 4F; aged 18–31.	Analysis and Methods: 2 types of whole brain analyses performed: VBM and a novel, quantile analysis. ROI analysis: Regions of interest were also analyzed using a quantile approach. Dependent Measures: FA, Trace. PCS, neurobehavioral battery, and quality of life assessments.	Whole brain analysis. 5 ROIs were selected and manually outlined in native space: anterior internal capsule, posterior internal capsule, anterior CC, posterior CC, and external capsule.	Whole brain analysis showed lower 1st percentile trace values in patients compared with controls. Trace from whole brain analysis correlated with PCS scores at 72 hrs and at 1 month post-injury. Trace from whole brain analysis was lower in the left anterior internal capsule, and showed higher median FA values in the posterior CC in patients compared with controls. FA values correlated with 72-h PCS score, visual motor speed, and impulse control. Findings suggest lower trace and increased FA may indicate axonal swelling early in the course of axonal injury, which correlates poorly with clinical outcome.
Kraus	2007	Chronic. ≥ 6 months out from injury; average 107 months for all TBI subjects.	3T	Patients: 37 TBI patients; 20 mTBI (8M, 12F; mean age 35.85) [LOC <30 min], 17 moderate to severe TBI (8M, 9F; mean age 34.88) [LOC >30 min and/or GCS <13]. 5 in each TBI group with other associated trauma. Controls: 18 controls (7M, 11F; mean age 32.83).	Analysis and Methods: ROI analysis. ROIs drawn on standardized space, FA maps. Dependent Measures: FA, RD, AD, white matter load defined as total number of regions with reduced FA. Neurocognitive measures relevant to attention, memory, and executive function.	13 ROI structures: anterior and posterior corona radiata, cortico-spinal tracts, cingulum fiber bundles, external capsule, forceps minor and major, genu, body and splenium of the CC, inferior fronto-occipital fasciculus, superior longitudinal fasciculus and sagittal stratum.	FA decreases were seen in all 13 ROIs for moderate/severe TBI, but only in the cortico-spinal tract, sagittal stratum, and superior longitudinal fasciculus in mTBI. Patients with moderate/severe TBI compared to controls showed increased AD and RD. Patients with mTBI compared to controls showed increases in AD, but not RD, suggesting myelin damage not present. White matter changes along a spectrum from mTBI to moderate/severe TBI and axonal but not myelin damage may characterize mTBI.
Lipton	2008	Chronic. 8 months to 3 years Retrospective study of consecutive admissions of mTBI.	1.5T	Patients: 17 mTBI patients (8M, 9F; age 26–70) [GCS 13–15, LOC <20 min]. Negative CT/MRI findings at time of initial injury. Later, at time of scanning, 1 mTBI subject had a small area of signal intensity likely due to gliosis. Controls: 10 controls of similar age and gender distribution. Negative MRI findings at time of scan.	Analysis and Methods: Whole-brain histogram analysis and Voxel- wise analysis of retrospective cases. Dependent Measures: FA, MD. Histogram parameters, i.e., kurtosis, skew.	Whole-brain analysis.	mTBI patients showed decreased FA and increased MD in CC, subcortical white matter, and internal capsules, bilaterally. Whole brain decrease in FA in mTBI by histogram analysis. Reduced FA and increased MD suggests that mTBI may be at one end of the diffuse axonal spectrum and the patients complaining of symptoms many months, and in some cases several years, post-injury are characterized by lack of white matter integrity in brain regions known to be affected in TBI.
Miles	2008	Acute. Average 4 days (range: 1–10). Follow-up at 6 months for neuropsychological measures.	1.5T	Patients: 17 patients (11M, 6F; mean age 33.44, range 18–58) [GCS 13–15, LOC <20 min]. Controls: 29 sex- and age-matched controls (15M, 14F; mean age 35, range 18–61). Not matched on education.	Analysis and Methods: ROI analysis. Dependent Measures: FA, MD. Neuropsychological measures at baseline (≤24 hour of imaging) and 6 month follow-up for 12/17 mTBI.	Structural ROI using circles: centra semiovale, the genu and the splenium of the CC, and the posterior limb of the internal capsule.	mTBI patients had increased MD and reduced FA compared with controls. Of the mTBI patients 41% had cognitive impairments at baseline and 33% at follow-up. At baseline no dependent measures were significantly correlated. Increased MD and decreased FA at baseline correlated with cognitive measures at follow-up. Findings suggest short-term predictors of cognitive function at 6 months. (Note: there was no DTI at 6 month follow-up).
Niogi	2008a	Subacute and Chronic. ≥ 1 month (range 1–65 months). ≥ 1 persistent symptom for postconcussive syndrome.	3T	Patients: 34 patients (18M, 16F; mean age 37.4, range 16–61). [GCS 13–15]. 11 mTBI- negative MRI findings; 11 mTBI micro- hemorrhage; 12 mTBI-non- specific white matter hyperintensities or chronic hemorrhagic contusions. Controls: 26 controls (19M, 7 7F; mean age 28.3, range 17–58).	Analysis and Methods: ROI Analysis. Manual ROI, ellipsoid shapes for each ROI. Dependent Measures: FA. Reaction time to Attention Network Task (Fan et al., 2005).	ROI structures: uncinate fasciculus, anterior corona radiata, anterior aspect of the inferior longitudinal fasciculus, genu of the CC, the cingulum bundle, and the superior longitudinal fasciculus.	Reduced FA in mTBI patients in anterior corona radiata (41% of patients), uncinate fasciculus (29%), genu of the CC (21%), inferior longitudinal fasciculus (21%), and cingulum bundle (18%). Reaction time on a simple cognitive task correlated to number of damaged white matter structures. (Note: 10/11 mTBI with negative MRI findings showed FA reductions relative to controls.)
Niogi (Extension of 2008a paper)	2008b	Subacute and Chronic. ≥1 month post-injury (mean 16.9 months, range 1–53 months). ≥ 1 persistent symptom for postconcussive syndrome.	3T	Patients: 43 patients (28M, 15F; mean age 32.4) [GCS 13–15]. Negative findings for conventional MRI images of 12/43 patients. Controls: 23 controls (17 M, 6F; mean age 29.9) Negative findings for conventional MRI.	Analysis and Methods: ROI analysis. Manual ROI, ellipsoid shapes for each ROI. Dependent Measures: FA. Attention Network Task (Fan et al., 2005), California Verbal Learning Test (CVLT) to test memory performance.	ROI structures: uncinate fasciculus, anterior corona radiata, anterior aspect of the inferior longitudinal fasciculus, genu of the CC, the cingulum bundle, and the superior longitudinal fasciculus.	mTBI and controls showed a significant correlation between attentional control, measured by the Attentional Network Task, and FA in left anterior corona radiata, as well as a significant correlation between memory performance, measured by the CVLT, and FA in the uncinate fasciculus, bilaterally. In mTBI, FA was reduced in both these brain regions compared with controls. Findings suggest FA can be used as a biomarker for neurocognitive function and dysfunction.
Rutgers	2008a	Subacute and Chronic. Median 5.5 months (range 0.1–109.3 months).	1.5T	Patients: 21 mTBI patients (12M, 9F; mean age 32±9) [GCS 13–15]. 17/21 patients negative findings on MRI; others contusions and 1 extra-axial hematoma. Controls: 11 controls (8M, 3F; mean age 37±9). No known history of positive MRI findings.	Analysis and Methods: ROI, tractography. Dependent Measures: FA, volume, ADC, number and length of through-passing fibers in each ROI, tractography measure of discontinuity.	ROI structures: cerebral lobar white matter, cingulum, CC, anterior and posterior limb of the internal capsules, mesencephalon, brain stem, and cerebellum. Cerebral lobar white matter was subdivided into centrum semiovale, frontal lobe, parietal lobe, temporal lobe, and occipital lobe.	On average, patients showed reduced FA in 9 regions, predominantly in cerebral lobar white matter, cingulum, and CC compared to controls (191 regions.) Discontinuity on fiber tracking was seen in only 19.3% of fibers. Other regions primarily involved supratentorial projection fiber bundles, callosal fibers, and fronto-temporo-occipital association fiber bundles. The internal capsules and infratentorial white matter were relatively unaffected. There was no association between time interval of scanning and post-injury measures.
Rutgers	2008b	Chronic. Average 2.8 months.	1.5T	Patients: 39 TBI patients (27M, 12F; mean age 34±12); 24 mild [GCS 13–15]; 9, moderate [GCS 9–12]; and 6, severe TBI [GCS ≤8]. Controls: 10 controls (7M, 3F; mean age 37±9). No known history of positive MRI findings.	Analysis and Methods: ROI analysis done by drawing ROI manually on the parts of the CC on FA maps of each subject. Dependent Measures: FA, ADC, number of fibers.	ROI structures: CC genu, body, and splenium.	Patients with mTBI < 3 months post-trauma (n=12) showed reduced FA and increased ADC in the genu of the CC. Patients with mTBI investigated > 3 months post-trauma (n=12) showed no differences. More severe trauma was associated with genu and splenium FA reduction, increased ADC, and fewer numbers of fibers.
Huang	2009	“Post-acute”.	1.5T	Patients: 10 patients mild TBI (mean age 25.0±11.5; mean education 12.7±4.7 years) [GCS 13–15]. 7 of 10 negative findings on conventional CT/MRI. Controls: 14 age-matched healthy subjects (mean age 27.4±15.2, mean education 12.9±3.2 years).	Analysis and Methods: Probabilistic tractography, Tract Based Spatial Statistics, MEG Dependent Measures: MEG delta slow waves, FA	Fronto-occipital fasciculus, superior longitudinal fasciculus, inferior longitudinal fasciculus, uncinate fasciculus, CC, and cingulum bundles. Automated Whole Brain Analysis.	Results presented as a case series. MEG slow waves originated from cortical gray areas with reduced DTI. In some cases, abnormal MEG delta waves were observed in subjects without obvious DTI abnormality, indicating that MEG may be more sensitive than DTI in diagnosing mTBI.
Lipton	2009	Acute. ≤ 2 weeks.	3T	Patients: 20 mTBI patients (9M, 11F; mean age 33.4) [CGS 13–15, LOC <20 min, PTA <24 hr]. Controls: 20 matched controls (9M, 11F; mean age 34.2).	Analysis and Methods: Voxel-wise analysis. Dependent Measures: FA, MD.	Automated analysis of whole brain.	Patients had multiple clusters of reduced FA in frontal white matter, including dorsolateral prefrontal cortex, with several clusters also showing higher MD as compared to controls. Reduced dorsolateral prefrontal cortex FA was significantly correlated with worse executive functioning in patients.
Lo	2009	Chronic. ≥ 2 years.	1.5T	Patients: 10 patients (5M, 5F; age range 20–51) [CGS 13–15]. For those who had conventional MR/CT, the findings were negative. Following research scan, 1 subject had a small focal area showing lobar gliosis. Controls: 10 controls (5M, 5F; mean age 44). Controls were patients referred for MRI due to headache and had no history of head trauma.	Analysis and Methods: ROI. Dependent Measures: FA, ADC.	ROIs placed in the genu and splenium of the CC, posterior limb of the internal capsule, and in the pontine tegmentum.	Compared to controls, patients had lower FA and higher ADC in the left genu of the CC. Patients showed increased FA in the posterior limb of the internal capsule bilaterally compared to controls.
Geary	2010	Chronic. ≥ 6 months.	3T	Patients: 40 mTBI patients (17M, 23F; mean age 34.53) [American Congress of Rehabilitation Medicine, 1993 criteria]. 14 subjects had previous head trauma. Negative MRI/CT findings. Controls: 35 controls (16M, 19F; mean age 32.54) matched to patients on age, education, years of employment, and estimated premorbid intelligence.	Analysis and Methods: ROI. Dependent Measures: FA.	ROIs structures: anterior and posterior corona radiata, corticospinal tracts (including parts of the corticopontine tract and superior thalamic radiation), external capsule, cingulum, forceps minor, forceps major, inferior fronto-occipital fasciculus, superior longitudinal fasciculus, uncinate fasciculus, sagittal stratum, and body, genu, and splenium of CC.	TBI patients had decreased FA in the superior longitudinal fasciculus, sagittal stratum, and uncinate fasciculus compared to controls. Poor performance on a verbal memory task was correlated with reduced FA in the uncinate fasciculus and in the superior longitudinal fasciculus in patients.
Holli(b)	2010	Acute. ≤ 3 weeks.	1.5T	Patients: 42 patients (17 M, 25F; mean age, 38.8) [GCS 13–15]. DTI analyses were performed on 34 patients. Negative CT/MRI findings. Controls: 10 age- and sex- matched controls (4M, 6F; mean age, 39.8).	Analysis and Methods: Texture analysis (MaZda). Dependent Measures: FA, ADC.	Regions corresponding to the mesencephalon, centrum semiovale, and CC.	Significant asymmetries in texture parameters were seen in all three regions studied in the patients, these asymmetries were less prevalent in the controls. Several texture parameters correlated with FA and ADC values. FA values in the left mesencephalon were positively correlated with verbal memory. ADC in the body of the CC was negatively correlated with visual memory.
Maruta	2010	Subacute to Chronic. 6 weeks and 5 years (mean 2.7 years).	3T	Patients: 17 patients (10M, 7F; age range 20–52), [GCS 15 with chronic postconcussive syndrome, recruited from clinics]. 8 normal, 9 showed abnormalities. Controls: 9 controls (6M, 3F; age range 19–31). No prior history of TBI.	Analysis and Methods: ROIs selected a priori. Dependent Measures: FA.	Anterior corona radiata, genu of the CC, uncinate fasciculus, cingulum bundle, forceps major, and superior cerebellar peduncle.	Gaze error variability correlated with the mean FA values of the right anterior corona radiata (ACR) and the left superior cerebellar peduncle, and the genu of the CC. Reaction time correlated to FA in the genu of the CC, and FA in the left uncinate fasciculus correlated with poorer verbal memory
Mayer	2010	Acute. ≤ 21 days (mean 12 days).	3T	Patients: 22 patients (mean age 27.45) [GCS 13–15, LOC <30 min, PTA <24hr]. Controls: 21 sex-, age-, and education-matched controls (mean age 26.81).	Analysis and Methods: ROI. Dependent Measures: FA, AD, RD.	Genu, splenium, and body of the CC, as well as the superior longitudinal fasciculus, the corona radiata, the superior corona radiata, the uncinate fasciculus, and the internal capsule for both hemispheres.	TBI patients had increased FA and reduced RD in the CC and several left hemisphere tracts compared to controls, i.e., uncinate fasciculus, internal capsule, and corona radiata. Follow-up data showed partial normalization of DTI values in several white matter tracts. The authors concluded that initial injury to axons might disrupt ionic hemeostasis and the balance of intra/extra cellular water, which effects diffusion that is perpendicular to the axons.
Zhang	2010	Subacute. 30 (±2) days.	3T	Patients: 15 student-athletes with mTBI (mean age 20.8 years) [Grade 1 MTBI according to Cantu 2006. GCS not reported]. Negative MRI findings. Controls: 15 normal student- athletes with no history of mTBI (mean age 21.3). Negative MRI findings. The entire sample was 70%M and 30%F.	Analysis and Methods: Voxel-wise whole brain analysis and ROI. Dependent Measures: FA, ADC.	The CC was chosen as a primary ROI and subdivided into the genu, body, and splenium. In addition, the right hippocampus, left and right dorsolateral prefrontal cortexes were evaluated.	Neither whole-brain analysis nor ROI analysis showed significant alteration in FA. mTBI patients showed larger variability of FA in the genu, and body of the CC than controls. mTBI patients showed decreased ADC in left and right dorsolateral prefrontal cortex compared to controls.
Warner	2010b	Acute. ≤1 week Follow-up at 4–16 months.	3T	Patients: 24 patients with diffuse traumatic axonal injury (16M, 8F; mean age 27.2) [mean GCS 6.4; range not reported]. Controls: None.	Analysis and Methods: Automated Segmentation and Volumetric Analysis (FreeSurfer). Dependent Measures: Regional Volumes, FA, MD.	CC, fornix body, bilateral fornix crus, bilateral perforant pathway, cingulum, uncinate fasciculus, and inferior fronto-occipital fasciculus.	The volume of the hippocampus correlated with the FA in the fornix body, fornix crus, perforant pathway, and cingulum bundle. Volume of the right precuneus correlated with MD in the fornix body and right uncinate fasciculus Note: Morphometric findings not related to DTI are presented in Table 2.
Bazarian	2011	Acute. (72 hours post-injury). Prospective cohort study. Subjects underwent DTI pre- and postseason within a 3-month interval. (Concussed subjects underwent repeat testing ≤ 72 hours of injury. Subjects who did not suffer a concussion during the season underwent repeat testing ≤ 1 week of the end of the sports season 3 months after initial testing).	3T	Patients: One athlete concussion [witnessed LOC >20 min, transient amnesia or confusion; GCS not reported] and 8 athletes with 26–399 subconcussive head blows. Controls: 6 control subjects. 2 control subjects had isolated minor orthopedic injuries.	Analysis and Methods: Whole brain analysis. FA and MD changes were measured in five ROIs. Dependent Measures: FA, MD.	External capsule, posterior and anterior CC, posterior and anterior limb of the internal capsule.	FA changes were highest for the concussion subject, intermediary for subjects with subconcussive head blows, and lowest for controls. MD changes were highest for the concussion subject, intermediary for subjects with subconcussive head blows and lowest for controls. The changes in FA and MD in the concussion subject were in the right corona radiata and right inferior longitudinal fasciculus. FA and MD changes from pre-season to post-season were in both directions, i.e., increased and decreased FA and MD, although more were in the expected direction, i.e., decreased FA and increased MD.
Cubon	2011	Subacute and Chronic. ≥1 month (mean 115 days, SD 104 days) for the sports-related concussion subjects. ≥ 1 year for subjects with moderate and severe TBI.	3T	Patients: 10 college students with concussion (5M, 5F; mean age 19.7 years) [Diagnosis based on International Consensus Agreements 2000, GCS not reported]. 2 moderate TBI subjects (mean age 20) [GCS 9–12]. 3 severe TBI (mean age 47.3) [GCS ≤8]. Controls: 10 sex- and age- matched athletes (5M, 5F; mean age 20.4 years). Sex- and age-matched controls, for the 2 moderate (mean age 22) and 3 severe TBI subjects (mean age 46).	Analysis and Methods: TBSS. Dependent Measures: FA, MD.	Whole brain WM skeleton (TBSS).	Concussed subjects showed increased MD in several clusters in the left hemisphere in comparison to controls. The largest cluster included sagittal stratum (including the inferior longitudinal fasciculus and inferior fronto-occipital fasciculus), retrolenticular part of the internal capsule, the posterior thalamic radiation (including the optic radiation), the acoustic radiation, and the superior longitudinal fasciculus. The second largest cluster was located along the superior longitudinal fasciculus, anterior and superior to the part of the superior longitudinal fasciculus.
Davenport	2011	Chronic. 2–5 years after blast injury.	3T	Patients: 25 veterans with mTBI (24M, 1F; mean age 36) [LOC <30 min, PTA <24 hours, and neurological symptoms; GCS not reported). Controls: 33 veterans (28 M, 5F; mean age 32.5).	Analysis and Methods: 20 standard probabilistic tractography-based ROIs. Dependent Measures: FA.	Forceps major, forceps minor, anterior thalamic radiations, hippocampal portion of cingulum, cingulum, corticospinal tract, inferior fronto-occipital fasciculus, inferior longitudinal fasciculus, superior longitudinal fasciculus, temporal portion of superior longitudinal fasciculus, uncinate.	A global pattern of lower white matter integrity was seen in blast mTBI patients but not in controls. Patients with multiple blast injuries tended to have a greater FA voxels decreased then individuals with a single blast injury.
Grossman	2011	Chronic. MRI and a neuropsychological battery either ≤ 1 year after injury, or >1 year after injury.	3T	Patients: 22 patients (14M, 8F; mean age 38.2) [GCS 13–15]. Controls: 14 healthy controls (9M, 5F; mean age 36.5) matched to patients according to gender, age, and formal education	Analysis and Methods: ROI. Dependent Measures: MK, FA, MD.	Thalamus and the anterior limb, genu, and posterior limb of the internal, the splenium of the CC, and the centrum semiovale.	Lowered DTI and DKI derived measures in the thalamus and the internal capsule were seen in patients examined one year after injury in comparison to controls. In addition to these regions, patients examined more than one year after injury also showed similar differences in the splenium of the corpus callosum and the centrum semiovale. Combined use of DTI and DKI provides a more sensitive tool for identifying brain injury. MK in the thalamus might be useful for early prediction of permanent brain damage and cognitive outcome.
Henry	2011	Acute. 1–6 days post-concussion. For patients, follow-up 6 months later; for controls, 2nd scan 18 months after initial scan.	3T	Patients: 18 athletes with mTBI (all M; mean age 22.08) [Diagnosis based on American Academy of Neurology Quality Standards Subcommittee and Neurology, 1997; GCS not reported]. Controls: 10 athletes (all M; mean age 22.81).	Analysis and Methods: Voxel-based approach (VBA). Dependent Measures: FA, MD, AD.	A voxel-wise 2 × 2 repeated-measures analysis of variance (ANOVA) using SPM8 to the derived scalar images (FA, AD, and MD) was used.	mTBI patients showed increased FA in dorsal corticospinal tracts and in the CC at both time points compared to control. mTBI patients showed increased AD in the right corticospinal tracts at both time points compared to controls. mTBI patients showed decreased MD in the corticospinal tracts and CC at both time points compared to controls.
Lange	2011	Subacute. 6–8 weeks.	3T	Patients: 60 mTBI patients (43M, 17F; mean age 30.8) [Closed head injury and; LOC >1minute, PTA >15 minutes, GCS <15, or abnormality on CT]. n=41 negative CT findings, N=15 positive CT findings, and n=4 not ordered. N=21 with postconcussive disorder based on ICD-10; N=39 without postconcussive disorder. Controls 34 controls (25M, 9F; mean age 37.1) with orthopedic/ soft- tissue injuries. N=0 positive findings, n=3 negative findings, and n=31 not ordered. 52.9% of controls met ICD-10 criteria for postconcussive disorder.	Analysis and Methods: ROIs. Dependent Measures: FA, MD.	Genu, body, and splenium of the CC.	No differences on FA or MD between mTBI patients and controls. mTBI patients reported more post-concussion symptoms than controls. mTBI patients showed a nonsignificant trend increase in MD in the splenium of the CC compared to controls.
MacDonald	2011	Subacute. ≤ 90 days.	1.5T	Patients: 63 patients (all M, median age 24) [Diagnosis based on U.S. military clinical criteria for traumatic brain injury; GCS not reported]. Controls: 21 controls (all M, median age 31) exposed to blasts, but none with sustained TBI.	Analysis and Methods: ROI. Dependent Measures: Relative anisotropy (RA), MD, AD, RD.	Genu and splenium of the CC, the right and left middle cerebellar peduncles, the right and left cerebral peduncles, the right and left uncinate fasciculi, and the right and left cingulum bundles.	TBI subjects showed decreased RA in the middle cerebellar peduncles, cingulum bundles, and in the right orbitofrontal white matter. Follow-up DTI scans in 47 TBI subjects 6–12 months later showed persistent abnormalities.
Matthews	2011	Self-reported history of concussion.	3T	Patients: 22 patients with a history of blast- related TBI. [LOC <30min, PTA <25 hr, CGS 13–15]. 11 TBI patients with Major Depressive Disorder (MDD) (all M, mean age 26.8). 11 TBI patients without MDD (all M, mean age 30.3). Controls: None.	Analysis and Methods: VBA. Dependent Measures: MD, 3 Lambda values, 9 eigenvalues λ1, λ2, λ3, total MD, FA.		Patients with MDD showed lower FA in corona radiata, CC and superior longitudinal fasciculus compared to non-MDD patients. FA in the superior longitudinal fasciculus correlated negatively with depressive symptoms.
Matsushita	2011	Acute. Median 3.5 days.	1.5T	Patients: 20 patients (18M, 2F; mean age 39.3): 9 mild TBI [GCS 13–15, LOC <30 min, PTA <24 hr]. 11 moderate TBI [LOC ≥30 min and/or GCS 9–12]. Controls: 27 matched controls (13M, 14F; mean age 42.9).	Analysis and Methods: ROI. Dependent Measures: FA.	Genu, stem, and splenium of the CC and the corona radiata, anterior limb of the internal capsule, posterior limb of the internal capsule, frontal white matter, and occipital white matter of the periventricular white matter.	Decreased FA in the splenium of the CC in mTBI compared to controls. Decreased FA in the genu, stem and splenium of the CC in moderate TBI compared to controls.
Messe	2011	Acute. 7–28 days (subacute phase), Subacute. 3–4 months (late phase).	1.5T	Patients: 23 patients [LOC <30 min, CGS 13–15, PTA <24 hr, or neurological symptoms]; 11 with good outcome (8M, 3F; mean age 27.8), 12 with poor outcome (7M, 5F; mean age 31.3). 7 of 11 with good outcome had negative MRI findings and 6 out of 12 poor outcome had negative MRI findings. Controls 23 controls (12M, 11F; mean age 30).	Analysis and Methods: VBM, TBSS. Dependent Measures: FA, MD, AD, RD.	Whole Brain.	MD was higher in poor outcome patients compared with both controls and good outcome patients, in the forceps major and minor of the CC, inferior fronto-occipital fasciculus, and inferior longitudinal fasciculus. MD was higher in poor outcome patients compared to controls in superior longitudinal fasciculus and corticospinal tract, and left anterior thalamic radiation. No difference in diffusion variables was found between patients with good outcome and controls.
Smits	2011	Subacute. Average 30.6 days.	3T	Patients: 19 patients included (10M, 9F; mean age 26) [GCS 13–15]. Controls: 12 controls (8M, 4F; mean age 28).	Analysis and Methods: TBSS. Dependent Measures: FA, MD.	Whole Brain.	There were no differences in MD between patients and controls. Decreased FA was in the right temporal subcortical white matter in patients compared to controls. These regions included inferior fronto-occipital fasciculus. Increased MD was correlated with severity of post-concussive symptoms in the inferior fronto-occipital fasciculus, inferior longitudinal fasciculus and the superior longitudinal fasciculus. Decreased FA correlated with severity of post-concussive symptoms in the uncinate fasciculus, inferior fronto-occipital fasciculus, internal capsule, CC, parietal subcortical white matter and frontal subcortical white matter. Microstructural changes in white matter are likely a neuropathological substrate of postconcussive syndrome.
Sponheim	2011	Chronic. Several Months.	1.5T	Patients: 9 patients (all M; mean age 33.7) [Diagnosis based on American Congress of Rehabilitation Medicine Special Interest Group on Mild Traumatic Brain Injury and the concussion grading system by the American Academy of Neurology; GCS not reported]. Controls: 8 controls (all M; mean age 30.3).	Analysis and Methods: ROI. Dependent Measures: Electroencephalogram (EEG) phase synchronization and FA.	Forceps major, forceps major and anterior thalamic radiations.	EEG phase synchrony was less in lateral frontal regions in patients. EEG phase synchrony was correlated with FA in the left anterior thalamic radiations and the forceps minor in patients but not controls.
McAllister	2012	Acute. Preseason and ≤ 10 days of a diagnosed concussion.	3T	Subjects: 10 athletes (all M; age range 15–23) with mTBI [Diagnosis by certified athletic trainer or team physician; GCS not reported]. Controls: None.	Analysis and Methods: Strain and strain rate calculations based on Dartmouth Subject- Specific FE Head model (helmets worn to record head impacts during play on the field). DTI maps for CC. Dependent Measures: FA, MD, Strain, Strain Rate.	CC.	Change in FA correlated with mean and maximum strain rate. Change in MD correlated with mean and maximum strain, Change in MD correlated with length of injury-to-imaging interval, but change in FA did not.

Key: M=Male; F=Female; PTA=Posttraumatic Amnesia; LOC=Loss of Consciousness; GCS=Glascow Coma Scale; CC=Corpus Callosum; ADC=Apparent Diffusion Coefficient; FA=Fractional Anisotropy; MD=Mean Diffusivity; RD=Radial Diffusivity; AD=Axial Diffusivity; DKI=Diffusional Kurtosis Imaging; DTI=Diffusion Tensor Imaging; MK=Mean Kurtosis; ROI=Region of Interest; TBSS=Tract-Based Spatial Statistics.

This table presents findings derived from diffusion tensor imaging (DTI) including: Trace, Mean Diffusivity (MD), Fractional Anisotropy (FA), Axial Diffusivity (AD), Radial Diffusivity (RD), and Mean Kurtosis (MK). In studies where non-DTI diffusion weighted imaging (DWI) was also presented we include the main measure, Apparent Diffusion Coefficient (ADC). In the description of subjects, we provide, when available, the Glasgow Coma Scale (GCS) score, Loss of Consciousness (LOC) duration, and Posttraumatic Amnesia (PTA) duration. Acute is defined as under 1 month, subacute is defined as >1 month but less than 6 months, and chronic is defined as >6 months.