Table 2.

MR Morphometry Studies

First Author	Year	Time Post-Injury	Magnet	Subjects	Analysis Method	Main Findings
Anderson	1995	Subacute. ≥ 6 weeks.	Not Specified	Patients: 68 TBI patients (49M, 19F), [GCS 3–15]; 38 with frontal lesions and 29 with no frontal lesions. Controls: None.	Volumetric Analysis.	No significant differences were observed between the two groups on lesion size and ventricle to brain ratio (VBR). No significant differences were observed between groups on tests of neuropsychological functioning.
Gale	1995	Chronic. Average 21 months.	1.5T	Patients: 88 TBI patients (51M, 37F; mean age 28.5), [GCS 3–15], 36 mTBI [GCS 11–15], 22 moderate TBI [GCS 7–10], 29 severe TBI [GCS 3–6]. Controls: 73 controls (36M, 37F; mean age 31).	Volumetric Analysis.	Patients had decreased fornix-to-brain ratio, total brain volume, CC volume; and increased temporal horn volume, ventricular volume, ventricle-to-brain ratio, and interpeduncular cistern volume. Fornix-to-brain ratio correlated with multiple neuropsychological measures. Cerebral peduncle and CC volume correlated with neuropsychological tests of motor function. GCS was correlated with fornix-to-brain ratio, internal capsule volume, brain volume, and cerebral peduncle volume.
Anderson	1996	Subacute and Chronic. ≥ 6 weeks.	Not Specified	Patients: 63 TBI patients (45M, 18F); 35 with lesions (26M, 9F; mean age 29.43), 28 without lesions (19M, 9F; mean age 31.33). Controls: 33 controls (26M, 9F; mean age 29.24).	Volumetric Analysis.	Subjects with lesions had significantly smaller thalami, higher VBR, and lower GCS. More severe injuries (GCS<9) had significantly smaller thalamic volumes and greater VBR than less severely injured subjects. VBR was negatively correlated with GCS and thalamic volume, and positively correlated with lesion volume. GCS was negatively correlated with lesion volume, and positively correlated with thalamic volume.
Bigler	1997	Subacute and Chronic. 44 scans ≤ 100 days post-injury, 55 scans > 100 days post-injury.	1.5	Patients: 94 TBI patients (59M and 35F; mean age 27) [GCS 3–15]. Controls: 96 controls (37M, 59F; mean age 31).	Volumetric Analysis.	Patients relative to controls showed reduced hippocampal volume and temporal horn enlargement. These measures correlated with cognitive outcome. In patients >70 days post-injury, temporal horn volume correlated with IQ, and hippocampal volume correlated with verbal memory function.
Tate	2000	Subacute and Chronic. ≥ 2 months.	1.5	Patients: 86 TBI patients (58M, 28F; mean age 30.9), [GCS 3–15]. Controls: 46 controls (31M, 15F; mean age 37.21).	Volumetric Analysis.	Decreased fornix and hippocampal volume in patients versus controls Fornix and hippocampal volumes correlated with degree of injury severity in patients. Hippocampal volume correlated with memory functioning.
MacKenzie	2002	Subacute and Chronic. 14 patients at >3 months after injury, 7 patients at 2 time points >3 months apart.	1.5T	Patients: 14 patients (mean age 36.1) [GCS 9–15]; 11 with mTBI [GCS 13–15], and 3 with moderate TBI [GCS 9–12]. Controls: 10 controls (mean age 34.9) underwent one MR session, and 4/10 underwent 2 sessions > 3 months apart.	Volumetric Analysis.	Brain volumes, CSF volumes, and % Volume of Brain Parenchyma (VBP) were not significantly different between patients and controls at the single time point. Rate of decline in %VBP and change in %VBP between the first and second time points were significantly greater in patients. Change in %VBP was greater in patients with loss of consciousness than in those without.
Yount	2002	Chronic. Average 22.8 months.	1.5T	Patients: 27 patients (18M, 9F; mean age 26) with TBI, [GCS 4–14]. No focal lesions or infarctions on MRI. Controls: 12 age and gender-matched controls.	Volumetric Analysis.	Patients had significant atrophy, primarily in the posterior cingulate gyrus, and degree of atrophy was correlated with severity of injury. Patients also had reduced CC and thalamic cross-sectional surface areas, with associated increased lateral ventricular volume, as well as reduced brain volume and increased ventricle-to-brain ratio. Neuropsychological performance was not related to changes in cingulate gyrus cross-sectional surface area in the TBI patients.
Bergeson	2004	Subacute and Chronic. ≥ 90 days.	1.5 T	Patients: 75 TBI patients (50M, 25 F; mean age 32.9) [GCS 3–14]. Controls: 75 controls (50M, 25F; mean age 31.4).	Volumetric Analysis.	Patients had significantly increased atrophy in frontal and temporal lobes compared to controls. Atrophy in frontal and temporal lobes correlated with deficits in memory and executive function.
Wilde	2004	Subacute and Chronic. ≥ 90 days.	1.5T	Patients: 77 patients [GCS 3–15]; 25 TBI and positive blood alcohol level (BAL) and 52 TBI with negative BAL. Controls: None.	Volumetric Analysis.	Increased general brain atrophy was observed in patients with a positive BAL and/or a history of moderate to heavy alcohol use.
Gale	2005	Chronic. Approximately 1 year (mean 10.6 months).	1.5T	Patients: 9 patients with a history of TBI (8M, 1F; mean age 29.1), [GCS 5–15]. Controls: 9 controls: (8M, 1F; mean age 28.8).	Voxel Based Morphometry.	Patients showed reduced gray matter concentration in frontal and temporal cortices, cingulate gyrus, subcortical gray matter, and in the cerebellum. Decreased gray matter concentration was also correlated with lower scores on tests of attention and lower GCS.
Himanen	2005	Chronic. Average 30 years.	1.5T	Patients: 61 patients (41M, 20F; mean age at injury 29.4); 17 mTBI [post-traumatic amnesia (PTA) <1 hr], 12 moderate TBI [PTA 1–24 hrs], 11 severe [PTA 1–7 days], 21 very severe [PTA >7 days] [GCS not reported]. Controls: None.	Volumetric Analysis.	Reduced hippocampal and increased lateral ventricular volumes were significantly associated with impaired memory functions, memory complaints, and executive functions. Volume of the lateral ventricles was the best predictor of cognitive outcome. There was also a modest relationship between severity of injury and cognitive performance.
Wilde	2006	Suacute and Chronic. ≥ 90 days.	1.5 T	Patients: 60 patients with severe-to-mild [GCS 3–15] TBI (38 M, 22 F; mean age 28.6). Controls: None.	Volumetric Analysis.	Longer Post-Traumatic Amnesia (PTA) duration predicted increased brain atrophy (ventricle to brain ratio). 6% increase in the odds of developing later atrophy with each additional day of PTA.
Cohen	2007	Acute to Chronic. 7 patients within 9 days, 13 patients from 1.2 months to 31.5 (average 4.6 years).	1.5T	Patients: 20 mTBI patients (11 M, 9F; median age 35), [GCS 13–15]. Controls: 19 controls (11M, 8F; median age 39).	Lesion Detection and Volumetric Analysis.	MR imaging-visible lesions were seen in 30% of patients. The most common lesions were punctate foci and gliosis distant from the trauma site, known to be associated with axonal injury and gliosis in regions susceptible to sheering injury. Patients exhibited increased global and gray matter atrophy. Brain atrophy did not differentiate patients with and without lesions.
Trivedi	2007	Subacute. Approximately 79 days; rescanned at approximately 409 days.	3T	Patients: 37 TBI patients (27M, 9F, 1 N/A; mean age 29.3); 11 mTBI [GCS ≥13], 10 moderate TBI [GCS 9–12], 16 severe TBI [GCS ≤8]. Controls: 30 controls (13M, 7F; mean age 24.5).	Volumetric Analysis.	Patients had a greater decline in % brain volume change (%BVC) compared to controls. Greater %BVC correlated with longer duration of coma post-injury.
Ding	2008	Acute and Chronic. (≤ 1 month) and (≥ 6 months).	3T	Patients: 20 patients with TBI (13M, 7F; mean age 26); 4 complicated-mild TBI [GCS ≥13], 4 moderate TBI [GCS 9–12]; and 12 severe TBI [GCS ≤8]. Controls: 20 controls (13M, 7F; mean age 28)	FLAIR Lesion Volume and Volumetric Analysis.	Change in whole brain, white matter, and gray matter volumes, correlated with acute FLAIR Lesions volume. Volume of acute FLAIR lesions correlated with volume of decreased FLAIR signal in the follow-up scans. Functional outcome correlated with acute FLAIR lesion volume and chronic cerebral atrophy. FLAIR lesions imaging were strongly predictive of post- traumatic cerebral atrophy.
Fujiwara	2008	Chronic. Approximately 1 year.	1.5T	Patients: 58 TBI patients underwent MRI; 12 mild [GCS 13–15], 27 moderate [GCS 9–12], and 19 severe [GCS 3–8]. 18 patients with focal cortical contusions and 40 with diffuse injury only. Controls: 25 controls; behavioral testing only.	Semi-Automated Segmentation (SABRE).	There were patterns of regional volume loss associated with test performance across all three behavioral measures. The tasks were sensitive to effects of TBI. Performance in the Smell Identification Test (SIT), Object Alternation (OA), and the Iowa Gambling Task (IGT) was correlated with gray matter loss including ventral frontal cortex. SIT was the most sensitive to ventral frontal cortex damage, even in patients without focal lesions. The SIT was further related to temporal lobe and posterior cingulate/retrosplenial volumes. OA and the IGT were associated with superior medial frontal volumes.
Levine	2008	Chronic. ≥ 1 year.	1.5T	Patients: 69 TBI patients; 13 mild [mean GCS 14.6], 30 moderate [mean GCS 11.1], 26 severe [mean GCS 6.7]. Controls: 12 age- and sex-matched controls.	Semi-Automated Segmentation (SABRE).	Stepwise dose response relationship between parenchymal volume loss and TBI severity. Patients with moderate and severe TBI were differentiated based on brain volume pattern from those with mild TBI, who were in turn differentiated from non-injured control subjects. Volume loss correlated with TBI severity, particularly widespread in white matter and sulcal/subdural CSF. The most reliable effects were observed in the frontal, temporal, and cingulate regions, although effects were observed to varying degrees in most brain regions. Focal lesions were associated with greater volume loss in frontal and temporal regions compared to subjects without focal lesions. Significantly reduced frontal and temporal volume loss seen in patients with diffuse injury compared to controls.
Schonberger	2009	Chronic. Average 2.3 years.	1.5T	Patients: 99 patients with mild to severe TBI (74 M, 24 F; mean age 34.5), [GCS 3–15]. Controls: None.	Lesion Volume, Automated Segmentation and Volumetric Analysis (SPM5).	Older age and longer post traumatic amnesia (PTA) correlated with larger lesion volumes in both gray and white matter in almost all brain regions. Older age was correlated with smaller gray matter volumes in most neo-cortical brain regions, while longer PTA was associated with smaller white matter volumes in most brain regions. Concluded that older age worsens the impact of brain injury; PTA is a good measure of severity but is not related to the particular kind of injury.
Holli	2010(a)	Acute. ≤ 3 weeks.	1.5T	Patients: 42 mTBI patients (17 M, 25F; mean age 38), [GCS 13–15]. All patients had normal clinical CT and MRI. Controls: 10 controls (4M, 6F; mean age 39.8).	Texture analysis (MaZda).	Significant differences were found in texture between the left and right side of mesencephalon, in the white matter of the corona radiate, and in portions of corpus callosum in patients. Controls were less asymmetrical in texture than patients.
Strangman	2010	Chronic. Average 11.5 years.	1.5T	Patients 50 TBI patients (36M, 14F; mean age 47.2) with TBI with reported memory difficulties; 12 mild [LOC ≤30 min, or GCS 13–15], 12 moderate [LOC 30 min −24 hr or GCS 9–12], 24 severe [LOC >24 hr or GCS <9], 2 n/a. Controls: None.	Automated Segmentation and Volumetric Analysis (FreeSurfer).	Volume reduction in the hippocampus, lateral prefrontal cortex, thalamus, and several subregions of the cingulate cortex predicted rehabilitation outcome.
Warner	2010a	Acute. Initial median 1 day post-injury. Follow-up at median 7.9 months post-injury.	3T	Patients: 25 patients with diffuse traumatic axonal injury (18M, 7F; mean age 26.8), [GCS 3–15]. Controls: 22 Controls (14M, 8F, mean age 32.4).	Automated Segmentation and Volumetric Analysis (FreeSurfer).	Patients showed major global brain atrophy with a mean whole-brain parenchymal volume loss of 4.5%. In patients significantly decreased volume was observed in the amygdala, hippocampus, thalamus, CC, putamen, precuneus, postcentral gyrus, paracentral lobule, and parietal and frontal cortices. Caudate and inferior temporal cortex did not show atrophy in patients. Loss of whole-brain parenchymal volume, atrophy of the inferior parietal cortex, pars orbitalis, pericalcarine cortex, and supramarginal gyrus predicted long-term disability.
Warner	2010b	Acute. Initial ≤ 1 week post-injury. Follow-up at 4–16 months post-injury.	3T	Patients: 24 patients with diffuse traumatic axonal injury (16M, 8F; mean age 27.2) [mean GCS 6.4; range not reported]. Controls: None.	Automated Segmentation and Volumetric Analysis (FreeSurfer).	Bilateral thalamus volumes correlated with processing speed. Amygdala, hippocampal, and a range of cortical region volumes were correlated with learning and memory performance. Thalamus, superior frontal, superior parietal, and precuneus cortices correlated with executive function. Subcortical and cortical brain volumes were associated with learning and memory and processing speed, but not executive function.
Yurgelun-Todd	2011	Not Reported.	3T	Patients: 15 patients with one or more TBI (all M; mean age 34.9); 14 with at least one mild TBI; 1 with moderate/severe TBI. [GCS not reported; (mild, moderate and severe according to The Ohio State University— TBI Identification Method (OSU-TBI)]. 14 subjects were normal on a clinical MRI. Controls 17 controls (all M; mean age 34.0).	Automated Segmentation and Volumetric Analysis (FreeSurfer).	Patients exhibited lower right, left and total frontal lobe volume than patients. Right and total cingulum FA correlated with suicidal ideation and impulsivity.

Key: M=male; F=female; GCS=Glascow Coma Scale; CC=Corpus Callosum. Acute defined as under 1 month, subacute is defined as >1 month but < 6 months, and chronic is defined as >6 months.