Upload
others
View
2
Download
0
Embed Size (px)
Citation preview
A review of magnetic resonance imaging and diffusion tensorimaging findings in mild traumatic brain injury
M. E. Shenton & H. M. Hamoda & J. S. Schneiderman & S. Bouix &
O. Pasternak & Y. Rathi & M.-A. Vu & M. P. Purohit & K. Helmer &
I. Koerte & A. P. Lin & C.-F. Westin & R. Kikinis & M. Kubicki &R. A. Stern & R. Zafonte
Published online: 22 March 2012
Abstract Mild traumatic brain injury (mTBI), also referredto as concussion, remains a controversial diagnosis becausethe brain often appears quite normal on conventional com-puted tomography (CT) and magnetic resonance imaging(MRI) scans. Such conventional tools, however, do notadequately depict brain injury in mTBI because they arenot sensitive to detecting diffuse axonal injuries (DAI), alsodescribed as traumatic axonal injuries (TAI), the major braininjuries in mTBI. Furthermore, for the 15 to 30 % of thosediagnosed with mTBI on the basis of cognitive and clinical
symptoms, i.e., the “miserable minority,” the cognitive andphysical symptoms do not resolve following the first3 months post-injury. Instead, they persist, and in somecases lead to long-term disability. The explanation givenfor these chronic symptoms, i.e., postconcussive syndrome,particularly in cases where there is no discernible radiologicalevidence for brain injury, has led some to posit a psychogenicorigin. Such attributions are made all the easier since bothposttraumatic stress disorder (PTSD) and depression are fre-quently co-morbid with mTBI. The challenge is thus to use
M. E. ShentonClinical Neuroscience Laboratory, Laboratory of Neuroscience,Department of Psychiatry, VA Boston Healthcare System,and Harvard Medical School,Brockton, MA, USA
M. E. Shenton (*) :H. M. Hamoda : J. S. Schneiderman :S. Bouix :O. Pasternak :Y. Rathi :M.-A. Vu :M. P. Purohit :I. Koerte :M. KubickiPsychiatry Neuroimaging Laboratory, Departments of Psychiatryand Radiology, Brigham and Women’s Hospital,and Harvard Medical School,1249 Boylston Street,Boston, MA 02215, USAe-mail: [email protected]: http://pnl.bwh.harvard.edu
H. M. HamodaDepartment of Psychiatry, Children’s Hospital, and HarvardMedical School,Boston, MA, USA
O. Pasternak :C.-F. WestinLaboratory of Mathematics in Imaging, Department of Radiology,Brigham and Women’s Hospital, and Harvard Medical School,Boston, MA, USA
M. P. Purohit : R. ZafonteSpaulding Rehabilitation Hospital,Massachusetts General Hospital and Harvard Medical School,Boston, MA, USA
M. P. PurohitDivision of Medicine, Beth Israel-Deaconness Medical Center,and Harvard Medical School,Boston, MA, USA
K. HelmerDepartment of Radiology, Athinoula A.Martinos Center,Massachusetts General Hospital, and Harvard Medical School,Charlestown, MA, USA
A. P. LinCenter for Clinical Spectroscopy, Department of Radiology,Brigham and Women’s Hospital, and Harvard Medical School,Boston, MA, USA
R. KikinisSurgical Planning Laboratory, MRI Division,Department of Radiology, Brigham and Women’s Hospital,and Harvard Medical School,Boston, MA, USA
R. A. SternCenter for The Study of Traumatic Encephalopathy,Departments of Neurology and Neurosurgery,Boston University Medical School,Boston, MA, USA
Brain Imaging and Behavior (2012) 6:137–192DOI 10.1007/s11682-012-9156-5
# Springer Science+Business Media, LLC (outside the USA) 2012
neuroimaging tools that are sensitive to DAI/TAI, such asdiffusion tensor imaging (DTI), in order to detect brain inju-ries in mTBI. Of note here, recent advances in neuroimagingtechniques, such as DTI, make it possible to characterize betterextant brain abnormalities in mTBI. These advances may leadto the development of biomarkers of injury, as well as tostaging of reorganization and reversal of white matter changesfollowing injury, and to the ability to track and to characterizechanges in brain injury over time. Such tools will likely beused in future research to evaluate treatment efficacy, giventheir enhanced sensitivity to alterations in the brain. In thisarticle we review the incidence of mTBI and the importance ofcharacterizing this patient population using objective radio-logical measures. Evidence is presented for detecting brainabnormalities in mTBI based on studies that use advancedneuroimaging techniques. Taken together, these findings sug-gest that more sensitive neuroimaging tools improve the de-tection of brain abnormalities (i.e., diagnosis) in mTBI. Thesetools will likely also provide important information relevant tooutcome (prognosis), as well as play an important role inlongitudinal studies that are needed to understand the dynamicnature of brain injury in mTBI. Additionally, summary tablesof MRI and DTI findings are included. We believe that theenhanced sensitivity of newer and more advanced neuroimag-ing techniques for identifying areas of brain damage in mTBIwill be important for documenting the biological basis ofpostconcussive symptoms, which are likely associated withsubtle brain alterations, alterations that have heretofore goneundetected due to the lack of sensitivity of earlier neuroimag-ing techniques. Nonetheless, it is noteworthy to point out thatdetecting brain abnormalities in mTBI does not mean thatother disorders of a more psychogenic origin are not co-morbid with mTBI and equally important to treat. They argu-ably are. The controversy of psychogenic versus physiogenic,however, is not productive because the psychogenic view doesnot carefully consider the limitations of conventional neuro-imaging techniques in detecting subtle brain injuries in mTBI,and the physiogenic view does not carefully consider the factthat PTSD and depression, and other co-morbid conditions,may be present in those suffering from mTBI. Finally, we endwith a discussion of future directions in research that will leadto the improved care of patients diagnosed with mTBI.
Keywords Mild traumatic braininjury . mTBI . TBI . Diffusion tensorimaging . DTI .Magnetic resonanceimaging .MRI . Diffusion-weightedimaging . DWI . Susceptibility-weightedimaging . SWI . Signature injuryof war . Concussion . Postconcussivesyndrome . Postconcussive symptoms . ComplicatedmTBI . UncomplicatedmTBI . Physiogenesis . Psychogenesis . Miserable minority
Introduction
The scope of the problem More than 1.7 million people eachyear in the United States experience a traumatic brain injury(TBI), with 75 to 85 % of these injuries categorized as mild(mTBI; CDC 2010; Faul et al. 2010; Bazarian et al. 2006).This number is likely an underestimate because it does notinclude those who are seen in private clinics or by primarycare physicians, nor does it include those who do not seekmedical treatment (Langlois et al. 2006). It is estimated, infact, that 14 % of mTBI patients are seen in private clinics orby their own doctors, with an additional 25 % receiving nomedical attention (Sosin et al 1996). Based on the largenumber of known and likely unknown cases, traumatic braininjury has been referred to as the “silent epidemic” (e.g.,Goldstein 1990). Recently, the public has become moreaware of TBI based on news reports of sports injuriesleading to long-term effects of repetitive trauma to the brain,as well as news reports about soldiers returning from Iraqand Afghanistan with TBI. With respect to the latter, themost frequent combat-related injury incurred by soldiersreturning from Iraq and Afghanistan is TBI, and most par-ticularly mTBI (Okie 2005). The frequency of these injurieshas led to TBI being called the “signature injury of war”(Okie 2005). Further, approximately 22 % of the woundedsoldiers arriving at Lundstuhl Regional Medical Center inGermany have head, neck, or face injuries, with cases ofTBI resulting primarily from improvised explosive devices(IEDs), landmines, high pressure waves from blasts, bluntforce injury to the head from objects in motion, and motorvehicle accidents (Okie 2005; Warden 2006). Of particularnote, mTBI characterizes most of the blast-induced traumaticbrain injuries seen in service members returning from Iraqand Afghanistan, with reports of 300,000 service memberssustaining at least one mTBI as of 2008 (Tanielian andJaycox 2008). Mild TBI is thus a major health problem thataffects both civilians and military populations. The estimatedeconomic cost is also enormous, with mTBI accounting for44 % of the 56 billion dollars spent annually in the UnitedStates in treating TBI (Thurman 2001).
Lack of radiological evidence Mild TBI is, however, diffi-cult to diagnose because often the brain appears quite nor-mal on conventional computed tomography (CT) andmagnetic resonance imaging (MRI) (e.g., Bazarian et al.2007; Inglese et al. 2005; Hughes et al. 2004; Iverson etal. 2000; Miller 1996; Mittl et al. 1994; Povlishock andCoburn 1989; Scheid et al. 2003). This lack of radiologicalevidence of brain injury in mTBI has led clinicians typicallyto diagnose mTBI on the basis of clinical and cognitivesymptoms, which are generally based on self-report, andare non-specific as they overlap with other diagnoses (e.g.,Hoge et al. 2008; Stein and McAllister 2009). To complicate
138 Brain Imaging and Behavior (2012) 6:137–192
matters further, while most of the symptoms in mTBI aretransient and resolve within days to weeks, approximately15 to 30 % of patients evince cognitive, physiological, andclinical symptoms that do not resolve 3 months post-injury(e.g., Alexander 1995; Bazarian et al. 1999; Bigler 2008;Rimel et al. 1981; Vanderploeg et al. 2007). Instead, thesesymptoms persist and in some cases lead to permanentdisability (Carroll et al. 2004a and b; Nolin and Heroux2006), and to what has been referred to as persistent post-concussive symptoms (PPCS), or postconcussive syndrome(PCS), although the latter term, “PCS,” is controversial(e.g., Arciniegas et al. 2005).
This “miserable minority” (Ruff et al. 1996) oftenexperience persistent postconcussive symptoms (PPCS) thatinclude dizziness, headache, irritability, fatigue, sleep distur-bances, nausea, blurred vision, hypersensitivity to light andnoise, depression, anxiety, as well as deficits in attention,concentration, memory, executive function, and speed of pro-cessing (e.g., Bigler 2008). Kurtzke and Kurland (1993) esti-mates the incidence of persistent symptoms as being equal tothe annual incidence of Parkinson’s disease, Multiple Sclero-sis, Guillain-Barré Syndrome, Motor Neuron Disease, andMyasthenia Gravis, combined. Moreover, the modal age forinjury is young, in the 20’s and 30’s. Thus mTBI affects alarge number of individuals in the prime of life, where there is,to date, no consistent or reliable correlations between cogni-tive/clinical symptoms and radiological evidence of braininjury based on conventional neuroimaging.
The explanation given for PPCS, particularly when there isno discernible radiological evidence, has led some to posit apsychogenic origin (e.g., Belanger et al. 2009; Hoge et al.2008; Lishman 1988; Machulda et al. 1988). More specifical-ly, Hoge and colleagues (2008; 2009) suggest that postcon-cussive symptoms reported by soldiers with mTBI are largelyor entirely mediated by posttraumatic stress disorder (PTSD)and depression. In their study, after controlling for both PTSDand depression, the only remaining symptom was headaches.Headaches, nonetheless, are an important symptom of TBI,particularly mTBI.
The term “miserable minority,” described above, has beenused to identify those who likely have a more psychogenicetiology to their symptoms (e.g., Ruff et al. 1996). Suchattributions are easy to make given that the symptoms ofmTBI, as noted above, overlap with other disorders (e.g.,Hoge et al. 2008). Belanger et al. (2009) also suggest thatmost of the symptoms reported by those with mTBI are likelythe result of emotional distress. Others have also argued thatemotional distress and/or psychiatric problems account forthose who continue to experience postconcussive symptoms(e.g., Belanger et al. 2009; Greiffenstein 2008; Hoge et al.2008; Lishman 1988; Machulda et al. 1988).
Persistent symptoms, however, may be the result of moresubtle neurological alterations that are beneath the threshold
of what can be detected using conventional neuroimagingtechniques that all too often do not reveal brain pathology inmTBI (e.g., Hayes and Dixon 1994; Huisman et al. 2004;Fitzgerald and Crosson 2011; Green et al. 2010; Miller 1996;Niogi and Mukherjee 2010). This is not at all surprising, sinceconventional techniques are not sensitive to detecting diffuse/traumatic axonal injuries (DAI/TAI), the major brain injuriesobserved in mTBI (e.g., Benson et al. 2007).
There is also evidence from the literature to suggest thatin several cases of mTBI where there was no radiologicalevidence of brain injury, autopsy following death frominjuries other than mTBI revealed microscopic diffuse axo-nal injuries that conventional neuroimaging tools did notdetect, presumably because they were not sufficiently sen-sitive (e.g., Adams et al. 1989; Bigler 2004; Blumbergs et al.1994; Oppenheimer 1968).
We would argue that the controversy between mTBIbeing psychogenic versus physiogenic in origin is not pro-ductive because the psychogenic view does not carefullyconsider the limitations of conventional neuroimaging tech-niques in detecting subtle brain injuries in mTBI, and thephysiogenic view does not carefully consider the fact thatPTSD and depression, and other co-morbid conditions, maybe present in those suffering from mTBI. Further, patientswith mTBI may complain more when their symptoms arenot validated. That is, when there is no radiological evi-dence that explains their symptoms, and yet they still expe-rience symptoms, these patients may complain morebecause of the lack of validation, versus those patientswho have radiological evidence that validates their symp-toms, leading them to complain less, simply because theyhave a medical explanation for their symptoms.
The challenge The challenge then is to use neuroimagingtools that are sensitive to DAI/TAI, such as Diffusion TensorImaging (DTI), to detect brain injuries in mTBI. Specifical-ly, with recent advances in imaging such as DTI it will nowbe possible to characterize better extant brain injuries inmTBI. Of note, DTI is a relatively new neuroimaging tech-nique that is sensitive to subtle changes in white matter fibertracts and is capable of revealing microstructural axonalinjuries (Basser et al. 1994; Pierpaoli and Basser 1996;Pierpaoli and Basser 1996), which are also potentially re-sponsible for persistent postconcussive symptoms.
Other promising techniques include susceptibility weight-ed imaging (SWI), which is sensitive to micro-hemorrhagesthat may occur in mTBI (e.g., Babikian et al. 2005; Haacke etal. 2004; Park et al. 2009; Scheid et al. 2007), and MagneticResonance Spectroscopy (MRS), which measures brainchemistry sensitive to neuronal injury and DAI (e.g., Babikianet al. 2006; Brooks et al. 2001; Garnett et al. 2000; Holshouseret al. 2005; Lin et al. 2005; Lin et al. 2010; Provencher 2001;Ross et al. 1998; Ross et al. 2005; Seeger et al. 2003; Shutter
Brain Imaging and Behavior (2012) 6:137–192 139
et al. 2004; Vagnozzi et al. 2010). In this review we focusprimarily on MRI and, most particularly, on DTI findings inmTBI. In a separate article in this special issue, Dr. AlexanderLin and colleagues review MRS, single photon emissiontomography (SPECT), and positron emission tomography(PET) findings relevant to brain chemistry alterations inmTBI, and Dr. Brenna McDonald and colleagues reviewfunctional MRI (fMRI) findings in mTBI. The reader is alsoreferred to Dr. Robert Stern and colleagues’ article, also in thisissue, which reviews the evidence for repetitive concussiveand subconcussive injuries in the etiology of chronic traumaticencephalopathy in sports-related injuries such as professionalfootball (see also Stern et al. 2011).
Focus of this review Here we present evidence for brainabnormalities in mTBI based on studies using advancedMRI/DTI neuroimaging techniques. Importantly, theseadvances make it possible to use more sensitive tools toinvestigate the more subtle brain alterations in mTBI. Theseadvances will likely lead to the development of biomarkersof injury, as well as to staging of reorganization and reversalof white matter and gray matter changes following injury,and to the ability to chart the progression of brain injuryover time. Such tools will also likely be used in futureresearch to evaluate treatment efficacy, given their enhancedsensitivity to alterations in the brain.
Taken together, the findings presented below suggestthat more sensitive neuroimaging tools improve the de-tection of brain injuries in mTBI (i.e., diagnosis). Thesetools will, in the near future, likely provide importantinformation relevant to outcome (prognosis), as well asplay a key role in longitudinal studies that are needed tounderstand the dynamic nature of brain injury in mTBI.We also believe that the enhanced sensitivity of newerand more advanced neuroimaging techniques for identi-fying brain pathology in mTBI will be important fordocumenting the biological basis of persistent postcon-cussive symptoms, which are likely associated withsubtle brain alterations, alterations that heretofore havegone undetected due to the lack of sensitivity of earlier,conventional neuroimaging techniques.
Below we provide a brief primer of neuroimaging tech-niques, although the reader is referred to Kou et al. (2010),Johnston et al. (2001), Le and Gean (2009), and Niogi andMukherjee 2010 for more detailed information. For a de-scription of the molecular pathophysiology of brain injury,the reader is referred to Barkhoudarian et al. (2011). Thereader is also referred to Dr. Erin Bigler’s article in thisspecial issue for information regarding post-mortem andhistological findings in mTBI as well as for a discussionof the physiological mechanisms underlying TBI. Dr. Bigleremphasizes that neuroimaging abnormalities are “gross indi-cators” of the underlying cellular damage resulting from
trauma-induced pathology. We concur and believe that wenow have neuroimaging tools that are sufficiently sensitiveto discern both more gross indicators of pathology, as wellas microstructural changes in white matter, and micro-hemorrhages using newer imaging technologies. The readeris also referred to Smith et al. (1995) and to several recent andexcellent reviews of neuroimaging findings in mTBI (e.g.,Belanger et al. 2007; Bruns and Jagoda 2009; Gentry 1994;Green et al. 2010; Hunter et al. 2011; Kou et al. 2010; Le andGean (2009); Maller et al. 2010; Niogi and Mukherjee 2010).Jang (2011) has also published a recent review of the use ofDTI in evaluating corticospinal tract injuries after TBI.
Following the brief primer, we present MRI and DTI find-ings relevant to mTBI. We used PUBMED to locate thesearticles. The following keywords were used: (MRI or DTI orDiffusion Tensor) AND (Concussion or Mild TBI or MildTraumatic Brain Injury or mTBI). The dates for the articlesselected were inclusive to September 16, 2011. We did notinclude articles that were case studies, nor did we includearticles that focused on pediatric and adolescent populations(see article in this special issue by Wilde and colleagues,which covers this topic). We also did not include articles thatdid not specify the severity of injury, but instead describedonly the mechanism of injury, i.e., falls, motor vehicle acci-dent, hit by tram (e.g., Liu et al. 1999). For the morphometricMRI empirical studies, we note that most included mild,moderate, and severe TBI, rather than mTBI alone. Conse-quently we included all three. This was less the case for theDTI empirical studies, where many focused only on mTBI.We were thus able to separate empirical studies that focusedsolely on mTBI from those that included several levels ofseverity, although we report on both. We include detailedsummary tables of MRI and DTI findings in order to providethe interested reader with a more in depth and detailed reviewof each empirical study included in this review. Following thereview of MRI and DTI findings, we present future directionsfor research in mTBI, which include the use of multiplemodalities for imaging the same patients, and the importanceof following patients longitudinally. We also present newimaging methods that go beyond advanced imagingapproaches reviewed here that, to date, are still as yet not usedroutinely in a clinical setting. The potential for developingbiomarkers to identify and to characterize mTBI is also pre-sented. The need here is critical as mTBI is not only difficult todetect but the injuries to the brain are heterogeneous, andbiomarkers are needed for individualized diagnosis as wellas for early and effective treatment interventions.
Neuroimaging primer and role of neuroimaging in mTBI
Overview TBI is a heterogeneous disorder and there is noone single imaging modality that is capable of characterizing
140 Brain Imaging and Behavior (2012) 6:137–192
the multifaceted nature of TBI. Advances in neuroimagingare, nonetheless, unprecedented and we are now able tovisualize and to quantify information about brain alterationsin the living brain in a manner that has previously not beenpossible. These advances began with computed axial tomog-raphy (CT) in the 1970’s, and then with magnetic resonanceimaging (MRI) in the mid-1980’s, with more refined andadvanced MR imaging over the last 25 years, including per-fusion weighted imaging (useful for measuring abnormalblood supply and perfusion), susceptibility-weighted imaging(SWI; useful for measuring micro-hemorrhages – e.g., Haackeet al. 2004; Park et al. 2009), magnetization transfer MRI(useful for measuring traumatic lesions – e.g., see review inLe and Gean 2009), diffusion weighted imaging (DWI; usefulfor measuring edema and developed initially for studies ofstroke – see review in Le and Gean (2009)), diffusion tensorimaging (DTI; useful for measuring white matter integrity –e.g., Basser et al. 1994), and functional MRI (fMRI; usefulfor measuring altered cortical responses to controlled stimuli –e.g., see article by McDonald et al. in this issue). Other neuro-imaging tools, although not a complete list, include positronemission tomography (PET; useful for measuring regionalbrain metabolism using 2-fluro-2-deoxy-d-glucose, bothhyper and hypo metabolism observed in TBI – see Le andGean (2009) for review), single photon emission tomography(SPECT; useful for measuring cerebral blood flow but lesssensitive to smaller lesions that are observed on MRI – seearticle by Lin et al. in this issue), and magnetic resonancespectroscopy (MRS; useful for measuring brain metabolites/altered brain chemistry – see article by Lin et al. in this issue).The clinical use of such tools lags behind their development,although the gap between development and clinical applica-tion is narrowing.
Below, we provide a brief primer for some of the neuro-imaging tools available today. We include skull films, CT,and MRI including DWI/DTI, and susceptibility weightedimaging. This primer is not detailed nor is it comprehensive.Instead, our intention is to provide the reader who is lessfamiliar with neuroimaging techniques with a context forsome of the tools available for investigating mTBI. Otherneuroimaging modalities, which will not be described here,include MRS, PET, SPECT, and fMRI. MRS, PET, andSPECT, will be reviewed by Dr. Alexander Lin and col-leagues, and Dr. Brenna McDonald and colleagues willreview fMRI, in separate articles in this issue. Table 1provides a brief summary of these neuroimaging tools.
Skull-X-ray and CT Skull films, or skull X-rays, whileexcellent for detecting skull fracture, are not used routinelyto investigate brain trauma because they provide very limit-ed information (e.g., Bell and Loop 1971; Hackney 1991).Figure 1 depicts a normal skull film. Computed Tomogra-phy or Computed Axial Tomography (CT) supplanted the
use of skull films for evaluating neurotrauma when thistechnology became available in the 1970s. CT providesthree-dimensional images of the inside of an object, in thiscase the brain, using two-dimensional X-Ray imagesobtained around a single axis of rotation. Since CT wasintroduced in the 1970s, it has become the imaging modalityof choice for evaluating closed head injury in the emergencyroom (ER) (e.g., Johnston et al. 2001).
CT is, in fact, the main imaging modality used in the first24 h for the management of neurotrauma in the ER (e.g.,Coles 2007). The reasons for this are because it is widelyavailable in most hospitals, it is fast, and it is accurate fordetecting emergent conditions such as skull fractures, brainswelling, intracranial hemorrhage, herniation, and radio-opaque foreign bodies in the brain (see review in Johnstonet al. 2001; Le and Gean (2009)). The use of thin-volumeCT scanners are also often located in close physical prox-imity to the ER, thus making it easy to transport neuro-trauma patients. Additionally, the presence of metallicobjects will not result in possibly dangerous accidents inthe CT suite as would be the case using an MR scan,depending upon the nature of the trauma, and dependingupon whether or not unknown small pieces of metal arehidden inside the patient following a car accident or othertype of brain trauma. Of further note, MRI scanners aregenerally not in close physical proximity to the ER, andthe scanning time is longer, which is an important consid-eration for patients who are not medically stable. Moreover,the CT environment is able to accommodate the set up of lifesupport and monitoring equipment that is, at this time, oftenmore compatible for the CT than for the MRI environment,although this is changing. CT thus remains the most impor-tant neuroimaging tool used in the first 24 h of acute neuro-trauma in the ER, where the most important question to beanswered quickly is: does this person need immediate neu-rosurgical intervention?
Figure 2 depicts a normal CT scan. Note that the skulland the brain are visible, although there is no differentiationbetween gray and white matter, which is discernible usingMRI. There are also bone artifacts with CT that are notpresent with MRI, which means that areas of injury aroundbone are easier to detect using MRI. MRI also uses noionizing energy, as CT does, which becomes importantwhen considering pediatric populations. This is also a con-sideration when several repeat scans are needed over time tofollow the progression of injury.
MRI and SWI Magnetic resonance imaging was introducedin the mid-1980s with the first images acquired on low-fieldmagnets, i.e., 0.5 Tesla (T). Originally this type of imagingwas called nuclear magnetic resonance (NMR) imaging butthe name was changed to magnetic resonance (MR) imag-ing, or MRI. The basic principle behind MRI is that
Brain Imaging and Behavior (2012) 6:137–192 141
radiofrequency (RF) pulses are used to excite hydrogennuclei (single proton) in water molecules in the human body,in this case the brain. By modulating the basic magneticfield, and the timing of a sequence of RF pulses, the scanner
produces a signal that is spatially encoded and results inimages. While NMR can be observed with a number ofnuclei, hydrogen imaging is the only one that is widely usedin the medical use of MRI.
MR images can be produced with different contrasts andcan be optimized to show excellent contrast between grayand white matter, which CT does not. Early MRI scans hadpoor spatial resolution and the time to acquire images wasslow, taking many minutes to acquire even one image. Sincethe mid-1980s, however, the field strengths of the magnethave increased from 0.5 to 1.0, to 1.5 T, and to 3.0 T andbeyond. In combination with advances in the capabilities ofthe gradient magnetic fields and the RF equipment available(parallel imaging), it is now possible to acquire sub-millimeter morphologic images and rich contrast combina-tions in clinical settings, in a shorter period of time. More-over, reconstruction algorithms can recreate images evenwhen the volume of the pixel elements (voxels) is notcompletely isotropic (i.e., the same size in all directions).Figure 3 depicts MRI scans acquired on a 3 T magnet using
Table 1 Summary of modalities
Imaging Technique/Modality: Function: Advantages Offered:
X-ray Imaging of bony structures Primarily used for detecting fractures.
Computed Tomography (CT) 3D X-ray imaging of an object (e.g., brainand skull).
Quick, able to have medical equipment inscanning area, good for skull fractures orgross injuries/abnormalities requiringemergent surgical intervention such assubdural hematomas.
Clinical Magnetic Resonance Imaging(MRI)
Uses radiofrequency pulses to detectchanges in spin signal of hydrogen atoms.
Better resolution than CT, particularly forsoft tissue, can provide gross delineationbetween gray and white matter structures,better visualization of brain stem areascompared to CT, can also detect subacutehemorrhages and macroscopic areas ofwhite matter damage.
Diffusion Weighted Imaging (DWI)/Diffusion Tensor Imaging (DTI)
Special type of MRI sequence that uses thediffusion properties of water to detectmicrostructural tissue architecture.
Best imaging technique available fordetecting white matter integrity/damage,able to detect microscopic white matterdamage and trace specific tracts of thebrain (e.g., corpus callosum, superiorlongitudinal fasciculus, uncinate).
Susceptibility Weighted Imaging (SWI) Special type of MRI technique that takesadvantage of susceptibility differencesamong structures (e.g., oxygenated vs.deoxygenated blood and iron).
Provides increased sensitivity to detect areasof micro-hemorrhage, particularly at gray-white matter junctions, that are not de-tectable on standard MRI.
Magnetic Resonance Spectroscopy (MRS) Measures brain chemistry by producing aspectrum where individual chemicals, ormetabolites can be identified andconcentrations can be measured.
Provides neurophysiological data that isrelated to structural damage/changes,neuronal health, neurotransmission,hypoxia, and other brain functions.
Positron Emission Tomography (PET) Uses radiotracers labeled with differentisotopes that emit signals indicating areasof uptake or binding in the brain, mostcommonly used is 18-Fluorodeoxyglucose, an analog of glucose.
Provides information on the concentrationof a chemical or protein in the brain suchas the amount of glucose, which reflectsactivity, or the density of a type of proteinsuch as beta amyloid, a hallmark ofneurodegenerative disease.
Fig. 1 Lateral (left) and frontal (right) view of normal skull X-ray.(Courtesy of Amir Arsalan Zamani, M.D.)
142 Brain Imaging and Behavior (2012) 6:137–192
1.5 mm slices. Note the high contrast between gray andwhite matter that is not visible on CT (see Figure 2). Cere-bral spinal fluid (CSF) is also prominent, and one can usethe differences in signal intensity of gray matter, whitematter, and CSF to parcellate automatically the brain intothese three tissue classes (e.g., Fischl et al. 2004; Pohl et al.2007). Of note here, the different tissue classes, from theparcellation, include quantitative information such as whole
brain volume for gray matter, white matter, and CSF. Thiswork is based on research developed over more than adecade in the field of computer vision.
Due to its superior contrast resolution for soft tissues,MRI technology is far more sensitive than CT in detectingsmall contusions, white matter shearing, small foci of axonalinjury, and small subacute hemorrhages (see review in Niogiand Mukherjee 2010). That MRI is able to discern these
Fig. 2 CT scan of a normalbrain. Left side is at the level ofthe temporal lobe where bonecan be seen as white areas(see red arrows). Right side is atthe level of the frontal lobe.(Courtesy of Amir ArsalanZamani, M.D.)
Fig. 3 Structural MRI scansacquired on a 3 T magnet using1.5 mm slices: a T1-weightedimage, b T2-weighted image,c T1-weighted image showinggray matter, white matter,and CSF parcellation, andd T1-weighted image showingthe corpus callosum regionof interest
Brain Imaging and Behavior (2012) 6:137–192 143
more subtle abnormalities, compared with CT, makes itparticularly well suited for evaluating mTBI. Additionally,there is higher contrast between brain and CSF, and betweengray and white matter, as well as better detection of edemawith MRI than CT, all important factors in evaluating TBI(see review in Johnston et al. 2001).
Of further note, and of particular interest to mTBI, Mittland coworkers (1994) found that in mTBI, where CT find-ings were negative, 30 % of these cases showed lesions onMRI that were compatible with hemorrhagic and non-hemorrhagic diffuse axonal injuries. The increased sensitiv-ity of MRI over CT in discerning radiological evidence ofbrain injury in mTBI has also been shown, and commentedupon, by a number of other investigators including Jenkinsand coworkers (1986), Levin and coworkers (1984; 1987),Eisenberg and Levin (1989), and Bazarian and coworkers(2007). Gentry and coworkers (1988) also observed that in aprospective study of 40 closed injury patients, MRI wassuperior to CT in detecting non-hemorrhagic lesions. Thesefindings, taken together, suggest that while CT may becritically important in the first 24 h to assess the immediateneed for neurosurgical intervention, for mTBI, MRI is likelyto be more sensitive for detecting small and subtle abnor-malities that are not detected using CT (e.g., Gentry et al.1988; Levin et al. 1987).
There are also several types of MRI sequences that add towhat can be gleaned from conventional MRI, including theuse of T1, T2-weighted FLAIR (FLuid Attenuated InversionRecovery) to examine macroscopic white matter lesions andcontusions on the cortical surface, as well as susceptibility-weighted imaging (SWI), which is a type of gradient-recalled echo (GRE) MRI that can be performed on conven-tional scanners. SWI was originally developed for venogra-phy and called Blood-Oxygen-Level-Dependent (BOLD)venographic imaging (Ashwal et al. 2006; Haacke et al.2009; Reichenbach et al. 2000; see also review in Kou etal. 2010 and Niogi and Mukherjee 2010). SWI takes advan-tage of susceptibility differences between tissues, resultingin an enhanced contrast that is sensitive to paramagneticproperties of intravascular deoxyhemoglobin, i.e., sensitiveto venous blood, to hemorrhage, and to iron in the brain. Inessence, susceptibility differences are detected as phasedifferences in the MRI signal. In the image processing stage,SWI superimposes these phase differences on the usual(magnitude) MR image, thereby allowing the susceptibilitydifferences to be accentuated in the final image. Of furthernote, SWI shows six times greater ability to detect hemor-rhagic diffuse axonal injuries than other MRI techniques(Tong et al. 2003; 2004). This technique is thus particularlyappropriate for discerning micro-hemorrhages in TBI, as itis sensitive to bleeding in gray/white matter boundaries,where small and subtle lesions are not discernible usingother MRI techniques, making it particularly useful in the
more acute and subacute stages following brain trauma.SWI, in conjunction with diffusion measures (e.g., DTI),will thus likely be important for discerning the subtle natureof mTBI abnormalities in the future. SWI is offered as alicensed acquisition and processing package by several ven-dors, but it can be acquired and processed on any scannersthat are 1.0 T, 1.5 T, 3.0 T, or above. Figure 4 depictssusceptibility-weighted images, where small black areasindicate blood vessels.
DWI and DTI Diffusion weighted imaging (DWI), devel-oped in 1991 for use in humans (e.g., Le Bihan 1991), isbased on the random motion of water molecules (i.e.,Brownian motion). This motion in the brain is affected bythe intrinsic speed of water displacement depending uponthe tissue properties and type, i.e., gray matter, white matter,and CSF. DWI was first used to evaluate acute cerebralischemia where it was thought that decreased diffusionwas the result of neuronal and glial swelling and likelyrelated to cytotoxic edema, whereas increased diffusionwas thought to reflect vasogenic edema. The method hasbeen applied to TBI with mixed results (see Niogi andMukherjee 2010).
Apparent Diffusion Coefficient (ADC) is a measure ofdiffusion, on average, and the word “apparent” is used toemphasize that what is quantified is at the level of the voxel,and not at the microscopic level. This measure has beenused as an indicator of edema, which, in conjunction withDTI (see below), can be used to quantify, indirectly, bothedema and damage to the integrity of white matter fiberbundles in TBI (see review in Assaf and Pasternak 2008;Niogi and Mukherjee 2010). A measure of free water, how-ever, derived from DTI (Pasternak et al. 2009; Pasternak etal. 2010; 2011a; b) may provide a better measure of edema,and this will be discussed further in the section on futuredirections of research.
DTI is a DWI technique that has opened up new possi-bilities for investigating white matter in vivo as it providesinformation about white matter anatomy that is not availableusing any other method — either in vivo or in vitro (Basseret al. 1994; Pierpaoli and Basser 1996; Pierpaoli and Basser1996; see also review in Assaf and Pasternak 2008). Attoday’s image resolution, it does not detect water behaviorwithin individual axons. Instead it describes local diffusionproperties. In other words, the individual behavior of axonscannot be described using DTI, but diffusion properties canbe described that are relevant to fiber bundles.
DTI differs from conventional MRI in that it is sensitiveto microstructural changes, particularly in white matter,whereas CT and conventional MRI (including also FLAIR)reveal only macroscopic changes in the brain. Thus subtlechanges using DTI can reveal microstructural axonal inju-ries (Basser et al. 1994; Pierpaoli and Basser 1996; Pierpaoli
144 Brain Imaging and Behavior (2012) 6:137–192
and Basser 1996), which are potentially responsible also forpersistent postconcussive symptoms.
The concept underlying DTI is that the local profile of thediffusion in different directions provides important indirectinformation about the microstructure of the underlying tis-sue. It has been invaluable in investigations of white matterpathology in multiple sclerosis, stroke, normal aging, Alz-heimer’s disease, schizophrenia and other psychiatric disor-ders, as well as in characterizing diffuse axonal injuries inmTBI (see reviews in Assaf and Pasternak 2008; Kou et al.2010; Shenton et al. 2010; Whitford et al. 2011).
The latter focus on TBI is relatively recent (see review ofthe literature, below). Those investigating mTBI, in partic-ular, have been disappointed by the lack of informationgleaned from conventional MRI and CT, although, as notedpreviously, this is not surprising given that the most com-mon injuries observed in mTBI are diffuse axonal injury/traumatic axonal injury (DAI/TAI), which are not easilydetected using conventional MR or CT scans. With theadvent of DTI, however, DAI/TAI have the potential to bequantified and this information can be used for diagnosis,prognosis, and for the evaluation of treatment efficacy.
Quantification of pathology using DTI is based on meas-ures that calculate the amount of restriction of water move-ment in the brain, which is determined to a large extent bythe tissue being measured. For example, the movement ofwater is unrestricted in a medium such as CSF, where itdiffuses equally in all directions (i.e., isotropic). However,in white matter, the movement of water is more restricted byaxonal membranes, myelin sheaths, microtubules, neurofila-ments, etc. In white matter, this restriction is dependent onthe directionality of the axons (i.e., diffusion is not equal inall directions) and is referred to as anisotropic diffusion.
Using tensors, adapted from the field of engineering, theaverage shape of the diffusion is characterized as more orless spherical when there is no impediment to water diffu-sion, as for example in CSF (i.e., unrestricted water is free todiffuse in all directions: isotropic). However, the average
shape of the diffusion becomes more elongated, or cigarshaped, when there is a preferred orientation in which wateris restricted, as for example in white matter. Here, waterdiffuses freely in directions parallel to axons but it is re-stricted in directions that are perpendicular to the axons,which results in the magnitude of the diffusion along theaxons being larger than the two perpendicular directions,leading to an elongated ellipsoidal shape of the diffusiontensor, described as anisotropic. The measurement of thedistance that water diffuses, over a given period of time, forat least six non-collinear directions, makes it possible toreconstruct a diffusion tensor (and the associated ellipsoid)that best describes water diffusion within a given voxel.Consequently, the volume (size) and shape of the ellipsoidcan be calculated, and this provides important informationabout the diffusion properties, and hence about microstruc-tural aspects of brain tissue.
There are various ways that the shape and size of adiffusion ellipsoid can be quantified, but the two mostcommon indices used are Fractional Anisotropy (FA) forshape, and Mean Diffusivity (MD) for size. FA is a scalarmeasure that ranges from 0 to 1, with 0 being completelyisotropic, meaning that water diffuses equally in all direc-tions, and 1 depicting the most extreme anisotropic scenarioin which molecules are diffusing along a single axis. Ac-cordingly, in CSF and gray matter, as noted above, thedirection of water is equal in all directions (i.e., isotropic),and the value is close to 0. In contrast, in white matter, forexample in the corpus callosum, the water is relatively freealong the axons, but restricted perpendicular to the axons,and therefore more anisotropic, with FA being closer to 1.Thus in white matter, reduced FA is generally thought toreflect loss of white matter integrity that may reflect damageto myelin or axon membrane damage, or perhaps reducedaxonal packing density, and/or reduced axonal coherence(see review in Kubicki et al. 2007).
Mean diffusivity (MD), the second most common mea-sure (and proportional to the trace of the diffusion tensor), is
Fig. 4 Sagittal (left) andaxial (right) view ofsusceptibility-weighted images(SWI) of a normal brain.Small black areas indicateblood vessels in the brainthat are enhanced using SWI
Brain Imaging and Behavior (2012) 6:137–192 145
different from FA in that it is a measure of the size of theellipsoid, rather than the shape, as is the case for FA. MD issimilar to ADC, described above for DWI, but instead it isthe average ADC along the three principal diffusion direc-tions, where one axis is in the direction of the largestmagnitude of the diffusion in the voxel, and the other twoare perpendicular to the main diffusion direction. The maindiffusion direction in white matter is referred to as thelongitudinal or axial direction, while the other two direc-tions are referred to as the radial or tangent axes. FA andMD are frequently observed as being inversely related. (Forfurther descriptions of DTI and associated methods of anal-yses, the reader is referred to Pierpaoli and Basser 1996;Pierpaoli and Basser 1996; Smith et al. 2006; and thereviews in Ashwal et al. 2006; Fitzgerald and Crosson2011; Hunter et al. 2011; Le and Gean 2009; Kou et al.2010; and Niogi and Mukherjee 2010).
Figure 5, 6, and 7 depict the kind information that can beextracted from diffusion tensor images. For example,Figure 5 shows diffusion images that highlight white matter,along with colored maps that reflect the directions of thewhite matter fiber tracts in the brain. Figure 6 shows whitematter tracts superimposed on structural images. Figure 7shows an area identified as tumor in the frontal lobe, wherewhite matter fiber tracts can be visualized in relation to thetumor and in relation to the frontal horn of the lateralventricles. These figures reflect important, recent advancesin methodology that are sufficiently robust and sensitive thatthey can be used for visualizing and quantifying whitematter pathology in vivo, for the assessment of mTBI clin-ically. These tools are available now for this purpose andwill be discussed further in the future directions section ofthis article.
DTI, however, is somewhat non-specific and it is notknown whether disruptions in FA and MD are the result ofdisturbances in axonal membranes, myelin sheath, micro-tubules, neurofilaments, or other factors. More specificmeasures, which are being developed (see below), are need-ed to delineate further the biological meaning of alterationsin white matter integrity (see review in Assaf and Pasternak2008; Niogi and Mukherjee 2010).
While FA and MD are the two main dependent measuresderived from DTI, there are other measures that have beendeveloped, including Mode (Ennis and Kindlmann 2006),which defines more precisely the shape of the diffusiontensor (useful in distinguishing the anatomy of fiber tracts,including distinguishing fiber crossings from pathology).Other measures include Inter-Voxel Coherence (Pfefferbaumet al. 2000), which measures how similar anisotropic tensorsare in neighboring voxels, useful in measuring anomalies inmacroscopic axonal organization within the tract of interest,and Axial and Radial Diffusivity, which are purported tomeasure axonal and myelin pathology, respectively (Song etal. 2001; Song et al. 2003; Budde et al. 2007; Budde et al.2011). These additional measures may provide more specificinformation regarding the microstructural abnormalitiesdiscerned using the sensitive, albeit less specific, measuresof FA and MD.
Finally, another relatively new post-processing method isfiber tractography, which was developed to visualize and toquantify white matter fiber bundles in the brain (e.g.,Conturo et al. 1999; Mori et al. 1999; Basser et al. 2000).This method makes it possible to follow fiber tracts along adiffusion direction in very small steps so as to create longfiber tracts that connect distant brain regions. The accuracyof fiber tractography is dependent upon a number of factorsincluding image resolution, noise, image distortions andpartial volume effects that result from multiple tracts cross-ing in a single voxel. The main advantage of DTI tractog-raphy, from a clinical research perspective, is that the wholefiber bundle, instead of just a portion of the fiber bundle, canbe evaluated. DTI tractography is thus a promising tool thatcan be used not only to understand how specific brainregions are connected and where damage occurs along fiberbundles, but it can also be used to understand how thisconnectivity may be relevant to functional abnormalities.Further, tractography methods can be used to both visualizeand to quantify white matter fiber bundle damage in a singlecase and thus these methods are potentially important fordiagnosing mTBI based on radiological evidence.
Importantly, many of the measures described above arejust beginning to be applied to investigate brain injuries in
Fig. 5 Diffusion tensor imagesacquired on a 3 T magnet. Left:fractional anisotropy (FA) map.White areas are areas of highanisotropy. Right: color byorientation map. Diffusion inthe left-right direction isshown in red, diffusion in thesuperior-inferior direction isshown in blue, and diffusion inthe anterior-posterior directionis shown in green
146 Brain Imaging and Behavior (2012) 6:137–192
Fig. 6 Fiber tractography ofcommonly damaged tracts inmild traumatic brain injury,including: a the anterior coronaradiata and the genu of corpuscallosum, b the uncinatefasciculus, c the cingulumbundle in green and the body ofcorpus callosum in red, andd the inferior longitudinalfasciculus (Niogi andMukherjee 2010; reprintedwith permission WoltersKluwer Health / LippincottWilliams & Wilkins)
Fig. 7 Diffusion MRI data forneurosurgical planning. Thetractography region of interest(ROI) is a box placed aroundthe tumor (in green) in thefrontal lobe. The ROI is alsovisualized with rectangles in theslice views below. Tracts arethen created based on theprincipal diffusion directions,which are color-coded (bottom).Diffusion ellipsoids are shownalong the tract to visualize theshape of the local diffusion
Brain Imaging and Behavior (2012) 6:137–192 147
mTBI and thus this area is a relatively new frontier forexploration. The application of DTI, and the measures de-rived from DTI, will likely contribute enormously to ourunderstanding of the nature and dynamics of brain injuriesin mTBI.
Review of MRI findings in mTBI
Much of the work with MRI has been to investigate thehigher sensitivity of MRI, compared with CT, for detectingbrain abnormalities in mTBI (see previous discussion). Lessattention has been given to investigating morphometricabnormalities in mTBI using area, cortical thickness, and/or volume measures. Table 2 lists studies, by first author andyear, which have examined aspects of morphometric abnor-malities in patients with mTBI. Most of these studies, how-ever, include a range of TBI, from mild to severe (e.g.,Anderson et al. 1995; Anderson et al. 1996; Bergeson etal. 2004; Bigler et al. 1997; Ding et al. 2008; Fujiwara et al.2008; Gale et al. 2005; Levine et al. 2008; Mackenzie et al.2002; Schonberger et al. 2009; Strangman et al. 2010; Tateand Bigler 2000; Trivedi et al. 2007; Warner et al. 2010a; b;Wilde et al. 2004; Wilde et al. 2006; Yount et al. 2002), withonly a small number of studies that investigate morphometricabnormalities specifically in mTBI (e.g., Cohen et al. 2007;Holli et al. 2010). Additionally, while most of the studieslisted in Table 2 categorize severity of TBI (i.e., mild, moder-ate, or severe) based on scores derived from the GlasgowComa Scale (GCS; Teasdale and Jennett 1974), one studydefines severity by posttraumatic amnesia (PTA) duration(Himanen et al. 2005).
The time of scan post-injury has also varied considerablyfrom study to study with the least amount of time being amedian of one day (Warner et al., 2010), up to a mean of30 years (Himanen et al. 2005), with one study that did notreport time of scan post-injury (Yurgelun-Todd et al. 2011).Additionally, most of these studies were performed using a1.5 T magnet, with only a small number performed using a3 T magnet (e.g., Ding et al. 2008; Trivedi et al. 2007;Warner et al. 2010a; b; Yurgelun-Todd et al. 2011). Thereare also different methods used to evaluate brain injuries,ranging from manual and automated measures of lesion vol-ume (e.g., Cohen et al. 2007; Ding et al. 2008; Schonberger etal. 2009), to volume analysis (e.g., Anderson et al. 1995;Anderson et al. 1996; Bergerson Bergeson et al. 2004; Bigleret al. 1997; Ding et al. 2008; Gale et al. 1995; Himanen et al.2005; Mackenzie et al. 2002; Schonberger et al. 2009;Strangman et al. 2010; Tate and Bigler 2000; Trivedi et al.2007; Warner et al. 2010a; b; Wilde et al. 2004; Wilde et al.2006; Yount et al. 2002), to voxel-based-morphometry(VBM; Gale et al. 2005), to texture analysis (Holli et al.2010a and b), to semi-automated brain region extraction
based template (SABRE) analysis (Fujiwara et al. 2008;Levine et al. 2008), to the use of FreeSurfer for volumetricanalysis of multiple brain regions (e.g., Strangman et al. 2010;Warner et al. 2010a; b; Yurgelun-Todd et al. 2011).
With all the differences among the studies, the mostimportant take home message is that MRI can be used todetect brain abnormalities in patients with TBI. It is also notsurprising that the injuries that are most apparent are ob-served in more moderate and severe cases of TBI. Further,the volume of lesions can be detected, although whether ornot these lesions are in frontal or non-frontal regions doesnot seem to differentiate between groups on measures ofcognitive function (Anderson et al. 1995). Mild TBIpatients, nonetheless, evince MR lesions in 30% of a sampleof 20 patients (Cohen et al. 2007), and in one study, func-tional outcome was correlated with lesion volume andcerebral atrophy, although this study did not analyze,separately, mild, moderate, and severe cases of TBI(Ding et al. 2008).
Overall brain volume reduction (atrophy) also seems tobe a common finding in what are likely to be more severepatients (e.g., Cohen et al. 2007; Ding et al. 2008; Gale et al.1995; Gale et al. 2005; Levine et al. 2008; Mackenzie et al.2002; Trivedi et al. 2007; Warner et al. 2010a; Yount et al.2002), and there are also volume reductions noted in overallgray matter (e.g., Cohen et al. 2007; Ding et al. 2008;Fujiwara et al. 2008; Schonberger et al. 2009; Trivedi etal. 2007), with a finding also of gray matter volume reduc-tion in the frontal lobe (e.g., Fujiwara et al. 2008; Strangmanet al. 2010; Yurgelun-Todd et al. 2011), and in frontal andtemporal lobes in some cases (e.g., Bergeson et al. 2004;Gale et al. 2005; Levine et al. 2008). Additionally, Bergesonet al. (2004) reported a correlation between frontal andtemporal lobe atrophy and deficits in memory and executivefunction in patients with a range of severity from mild, tosevere (GCS; 3-14).
Overall reduction in white matter has also been reported(e.g., Ding et al. 2008; Levine et al. 2008; Schonberger et al.2009), as well as white matter reduction at the level of themesencephalon, corona radiata, centrum semiovale (Holli etal. 2010a and b), and corpus callosum (Holli et al. 2010aand b; Warner et al. 2010a; Yount et al. 2002). Ding andcoworkers noted that the changes in white and gray matterover time were correlated with acute diffuse axonal injuriesand the latter predicted post-injury cerebral atrophy.
More specific reductions in volume in brain regions havealso been observed, including in the hippocampus (Bigler etal. 1997; Himanen et al. 2005; Strangman et al. 2010; Tateand Bigler 2000; Warner et al. 2010a), amygdala (e.g.,Warner et al. 2010a; b), fornix (Gale et al. 1995; Tate andBigler 2000), thalamus (e.g., Strangman et al. 2010; Warneret al. 2010a; Yount et al. 2002), regions involving thecingulate gyrus (e.g., Gale et al. 2005; Levine et al. 2008;
148 Brain Imaging and Behavior (2012) 6:137–192
Tab
le2
MRmorph
ometry
stud
ies
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
And
erson
1995
Sub
acute.
≥6weeks.
Not
Specified
Patients:68
TBIpatients(49M,
19F),[G
CS3-15
];38
with
fron
tal
lesion
sand29
with
nofron
tal
lesion
s.
Volum
etricAnalysis.
-Nosign
ificantdifferenceswere
observed
betweenthetwogrou
pson
lesion
size
andventricleto
brain
ratio
(VBR).
Con
trols:Non
e.-Nosign
ificantdifferenceswere
observed
betweengrou
pson
testsof
neurop
sycholog
ical
functio
ning
.
Gale
1995
Chron
ic.
1.5T
Patients:88
TBIpatients(51M,
37F;meanage28
.5),[G
CS3-
15],36
mTBI[G
CS11-15],22
mod
erateTBI[G
CS7-10
],29
severe
TBI[G
CS3-6].
Volum
etricAnalysis.
-Patientshaddecreasedfornix-to-
brainratio
,totalbrainvo
lume,CC
volume;
andincreasedtempo
ral
horn
volume,ventricularvo
lume,
ventricle-to-brain
ratio
,and
interpedun
cularcisternvo
lume.
Average
21mon
ths.
Con
trols:73
controls(36M,37
F;
meanage31
).-Fornix-to-brain
ratio
correlated
with
multip
leneurop
sycholog
ical
measures.
-Cerebralpedu
ncle
andCCvo
lume
correlated
with
neurop
sycholog
ical
testsof
motor
functio
n.
-GCSwas
correlated
with
fornix-to-
brainratio
,internal
capsulevo
lume,
brainvo
lume,andcerebralpedu
ncle
volume.
And
erson
1996
Sub
acuteandChron
ic.
Not
Specified
Patients:63
TBIpatients(45M,
18F);35
with
lesion
s(26M,9
F;
meanage29
.43),28
with
out
lesion
s(19M,9F;meanage
31.33).
Volum
etricAnalysis
-Sub
jectswith
lesion
shad
sign
ificantly
smallerthalam
i,high
erVBR,andlower
GCS.
≥6weeks.
-Moresevere
injuries
(GCS<9)
had
sign
ificantly
smallerthalam
icvo
lumes
andgreaterVBRthan
less
severely
injuredsubjects.
Con
trols:33
controls(26M,9F;
meanage29
.24).
-VBRwas
negativ
elycorrelated
with
GCSandthalam
icvo
lume,and
positiv
elycorrelated
with
lesion
volume.
-GCSwas
negativ
elycorrelated
with
lesion
volume,andpo
sitiv
ely
correlated
with
thalam
icvo
lume.
Bigler
1997
Sub
acuteandChron
ic.
1.5
Patients:94
TBIpatients(59M
and
35F;meanage27
)[G
CS3-15
].Volum
etricAnalysis.
-Patientsrelativ
eto
controlsshow
edredu
cedhipp
ocam
palvo
lumeand
44scans≤
100days
post-injury,
55scans>
100days
post-injury.
Brain Imaging and Behavior (2012) 6:137–192 149
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
temporalhorn
enlargem
ent.These
measurescorrelated
with
cogn
itive
outcom
e.
Con
trols:96
controls(37M,59
F;
meanage31
).-In
patients>70days
post-injury,
temporalhornvolumecorrelated
with
IQ,and
hippocam
palvolum
ecorrelated
with
verbalmem
oryfunction.
Tate
2000
Sub
acuteandChron
ic.
≥2mon
ths.
1.5
Patients:86
TBIpatients(58M,
28F;
meanage30.9),[G
CS3-15].
Volum
etricAnalysis.
-Decreased
fornix
andhipp
ocam
pal
volumein
patientsversus
controls
Con
trols:46
controls(31M,15
F;
meanage37
.21).
-Fornixandhipp
ocam
palvo
lumes
correlated
with
degree
ofinjury
severity
inpatients.
-Hippo
campalv
olum
ecorrelated
with
mem
oryfunctio
ning
.
MacKenzie
2002
Sub
acuteandChron
ic.
1.5T
Patients:14
patients(m
eanage
36.1)[G
CS9-15
];11
with
mTBI
[GCS13
-15],and3with
mod
erateTBI[G
CS9-12
].
Volum
etricAnalysis.
-Brain
volumes,C
SFvo
lumes,and
%Volum
eof
Brain
Parenchym
a(V
BP)
wereno
tsign
ificantly
different
betweenpatientsandcontrolsat
the
sing
letim
epo
int.
14patientsat
>3mon
thsafter
injury,7patientsat
2tim
epo
ints>3mon
thsapart.
Con
trols:10
controls(m
eanage
34.9)un
derw
enton
eMRsession,
and4/10
underw
ent2sessions>
3mon
thsapart.
-Rateof
declinein
%VBPandchange
in%VBPbetweenthefirstand
second
timepo
intswere
sign
ificantly
greaterin
patients.
-Chang
ein
%VBPwas
greaterin
patientswith
loss
ofconsciou
sness
than
inthosewith
out.
You
nt20
02Chron
ic.
1.5T
Patients:27
patients(18M,9F;
meanage26
)with
TBI,
[GCS4-14
].Nofocallesion
sor
infarctio
nson
MRI.
Volum
etricAnalysis.
-Patientshadsign
ificantatroph
y,prim
arily
inthepo
steriorcing
ulate
gyrus,anddegree
ofatroph
ywas
correlated
with
severity
ofinjury.
Average
22.8
mon
ths.
Con
trols:12
ageandgend
er-
matched
controls.
-Patientsalso
hadredu
cedCCand
thalam
iccross-sectionalsurface
areas,with
associated
increasedlat-
eral
ventricularvo
lume,as
wellas
redu
cedbrainvo
lumeandincreased
ventricle-to-brain
ratio
.
-Neuropsycho
logicalperformance
was
notrelatedto
changesin
150 Brain Imaging and Behavior (2012) 6:137–192
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
cing
ulategy
ruscross-sectional
surfacearea
intheTBIpatients.
Bergeson
2004
Sub
acuteandChron
ic.
1.5T
Patients:75
TBIpatients(50M,
25F;m
eanage32
.9)[G
CS3-14
].Volum
etricAnalysis.
-Patientshadsign
ificantly
increased
atroph
yinfron
taland
tempo
rallob
escomparedto
controls.
≥90
days.
Con
trols:75
controls(50M,25
F;
meanage31
.4).
-Atrop
hyin
fron
talandtempo
ral
lobescorrelated
with
deficitsin
mem
oryandexecutivefunctio
n.
Wild
e20
04Sub
acuteandChron
ic.
1.5T
Patients:77
patients[G
CS3-15
];25
TBIandpo
sitiv
ebloo
dalcoho
llevel(BAL)and52
TBIwith
negativ
eBAL.
Volum
etricAnalysis.
-Increasedgeneralbrainatroph
ywas
observed
inpatientswith
apo
sitiv
eBALand/or
ahistoryof
mod
erateto
heavyalcoho
luse.
≥90
days.
Con
trols:Non
e.
Gale
2005
Chron
ic.
1.5T
Patients:9patientswith
ahistoryof
TBI(8
M,1F;meanage29
.1),
[GCS5-15
].
Vox
elBased
Morph
ometry.
-Patientsshow
edredu
cedgray
matter
concentrationin
fron
taland
tempo
ralcortices,cing
ulategy
rus,
subcortical
gray
matter,andin
the
cerebellu
m.
App
roximately1year
(mean10
.6mon
ths).
Con
trols:9controls:(8
M,1F;
meanage28
.8).
-Decreased
gray
matterconcentration
was
also
correlated
with
lower
scores
ontestsof
attentionandlower
GCS.
Him
anen
2005
Chron
ic.
1.5T
Patients:61
patients(41M,20
F;
meanageat
injury
29.4);17
mTBI[post-traumatic
amnesia
(PTA
)<1hr],12
mod
erateTBI
[PTA
1–24
hrs],11
severe
[PTA
1-7days],21
very
severe
[PTA
>7days][G
CSno
trepo
rted].
Volum
etricAnalysis.
-Reduced
hipp
ocam
palandincreased
lateralventricularvo
lumes
were
significantly
associated
with
impaired
mem
oryfunctio
ns,m
emory
complaints,andexecutive
functio
ns.
Average
30years.
Con
trols:Non
e.-Volum
eof
thelateralventricles
was
thebestpredictorof
cogn
itive
outcom
e.
-There
was
also
amod
estrelationship
betweenseverity
ofinjury
and
cogn
itive
performance.
Wild
e20
06Suacute
andChron
ic.
1.5T
Patients:60
patientswith
severe-
to-m
ild[G
CS3–15
]TBI(38M,
22F;meanage28
.6).
Volum
etricAnalysis.
-LongerPost-Traum
aticAmnesia
(PTA
)duratio
npredictedincreased
brainatrophy(ventricletobrainratio
).≥90
days.
Con
trols:Non
e.-6%
increase
intheod
dsof
developing
lateratroph
ywith
each
additio
naldayof
PTA
.
Brain Imaging and Behavior (2012) 6:137–192 151
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
Coh
en20
07Acute
toChron
ic.
1.5T
Patients:20
mTBIpatients(11M,
9F;m
edianage35),[G
CS13-15].
LesionDetectio
nand
Volum
etricAnalysis.
-MRim
aging-visiblelesion
swere
seen
in30
%of
patients.
7patientswith
in9days,
13patientsfrom
1.2mon
thsto
31.5
(average
4.6years).
Con
trols:19
controls(11M,8F;
medianage39
).-The
mostcommon
lesion
swere
punctatefociandgliosisdistantfrom
thetraumasite,kn
ownto
beassociated
with
axon
alinjury
and
gliosisin
region
ssusceptib
leto
sheering
injury.
-Patientsexhibitedincreasedglob
alandgray
matteratroph
y.
-Brain
atroph
ydidno
tdifferentiate
patientswith
andwith
outlesion
s.
Trivedi
2007
Sub
acute.
3T
Patients:37
TBIpatients(27M,
9F,
1N/A;meanage29
.3);11
mTBI[G
CS≥1
3],10
mod
erate
TBI[G
CS9-12
],16
severe
TBI
[GCS≤8
].
Volum
etricAnalysis.
-Patientshadagreaterdeclinein
%brainvo
lumechange
(%BVC)
comparedto
controls.
App
roximately79
days;
rescannedat
approx
imately
409days.
-Greater
%BVCcorrelated
with
long
erdu
ratio
nof
comapo
st-injury.
Con
trols:30
controls(13M,7F;
meanage24
.5).
Ding
2008
Acute
andChron
ic.
3T
Patients:20
patientswith
TBI
(13M,7F;meanage26
);4
complicated-m
ildTBI[G
CS
≥13],4mod
erateTBI[G
CS9-
12];and12
severe
TBI[G
CS≤8
].
FLAIR
LesionVolum
eand
Volum
etricAnalysis.
-Chang
ein
who
lebrain,
whitematter,
andgray
mattervo
lumes,correlated
with
acuteFLAIR
Lesions
volume.
(≤1mon
th)and(≥
6mon
ths).
-Volum
eof
acuteFL
AIR
lesions
correlated
with
volumeof
decreased
FLAIR
signalin
thefollo
w-upscans.
Con
trols:20
controls(13M,7F;
meanage28
)-Fun
ctionalou
tcom
ecorrelated
with
acuteFLAIR
lesion
volumeand
chroniccerebral
atroph
y.
-FLAIR
lesion
sim
agingwere
strong
lypredictiv
eof
post-traum
atic
cerebral
atroph
y.
Fujiwara
2008
Chron
ic.
1.5T
Patients:58
TBIpatientsun
derw
ent
MRI;12
mild
[GCS13
-15],27
mod
erate[G
CS9-12
],and19
se-
vere
[GCS3-8].18
patientswith
focalcortical
contusions
and40
with
diffuseinjury
only.
Sem
i-Autom
ated
Segmentatio
n(SABRE).
-There
werepatternsof
region
alvo
lumeloss
associated
with
test
performance
across
allthree
behavioralmeasures.The
taskswere
sensitive
toeffectsof
TBI.
App
roximately1year.
Con
trols:25
controls;behavioral
testingon
ly.
-Perform
ance
intheSmell
IdentificationTest(SIT),Object
Alternation(O
A),andtheIowa
Gam
blingTask
(IGT)was
correlated
with
gray
matterloss
includ
ing
152 Brain Imaging and Behavior (2012) 6:137–192
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
ventralfron
talcortex.SIT
was
the
mostsensitive
toventralfron
tal
cortex
damage,even
inpatients
with
outfocallesion
s.
-The
SIT
was
furtherrelatedto
tempo
rallobe
andpo
sterior
cing
ulate/retrosplenialvo
lumes.
-OAandtheIG
Twereassociated
with
superior
medialfron
talvo
lumes.
Levine
2008
Chron
ic.
1.5T
Patients:69
TBIpatients;13
mild
[meanGCS14
.6],30
mod
erate
[meanGCS11.1],26
severe
[meanGCS6.7].
Sem
i-Autom
ated
Segmentatio
n(SABRE).
-Stepw
isedo
se–respon
serelatio
nship
betweenparenchy
mal
volumeloss
andTBIseverity.
≥1year.
Con
trols:12
age-
andsex-matched
controls.
-Patientswith
mod
erateandsevere
TBIweredifferentiatedbasedon
brainvo
lumepattern
from
those
with
mild
TBI,who
werein
turn
differentiatedfrom
non-injured
controlsubjects.
-Volum
eloss
correlated
with
TBI
severity,particularly
widespreadin
white
matterandsulcal/sub
dural
CSF.
-The
mostreliableeffectswere
observed
inthefron
tal,tempo
ral,
andcing
ulateregion
s,althou
gheffectswereob
served
tovarying
degreesin
mostbrainregion
s.
-Focal
lesion
swereassociated
with
greatervo
lumeloss
infron
taland
tempo
ralregion
scomparedto
subjectswith
outfocallesion
s.
-Significantly
redu
cedfron
taland
tempo
ralvo
lumeloss
seen
inpatientswith
diffuseinjury
comparedto
controls.
Schon
berger
2009
Chron
ic.
1.5T
Patients:99
patientswith
mild
tosevereTBI(74M,24F;m
eanage
34.5),[G
CS3-15
].
LesionVolum
e,Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(SPM5).
-Older
ageandlonger
posttraumatic
amnesia(PTA
)correlated
with
larger
lesion
volumes
inbothgray
andwhite
matterin
almostallbrainregions.
Average
2.3years.
Con
trols:Non
e.-Older
agewas
correlated
with
smallergray
mattervo
lumes
inmost
Brain Imaging and Behavior (2012) 6:137–192 153
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
neo-cortical
brainregion
s,while
long
erPTA
was
associated
with
smallerwhite
mattervo
lumes
inmostbrainregion
s.
-Con
clud
edthat
olderageworsens
theim
pact
ofbraininjury;PTA
isa
good
measure
ofseverity
butisno
trelatedto
theparticular
kind
ofinjury.
Holli
2010
(a)
Acute.
1.5T
Patients:42
mTBIpatients(17M,
25F;m
eanage38
),[G
CS13
-15].
Allpatientshadno
rmal
clinical
CTandMRI.
Texture
analysis(M
aZda).
-Significant
differenceswerefoundin
texturebetweentheleftandrightside
ofmesencephalon,inthewhitematter
ofthecorona
radiate,andin
portions
ofcorpus
callo
sum
inpatients.
≤3weeks.
Con
trols:10
controls(4
M,6F;
meanage39
.8).
-Con
trolswereless
asym
metricalin
texturethan
patients.
Strangm
an20
10Chron
ic.
1.5T
Patients50
TBIpatients(36M,
14F;meanage47
.2)with
TBI
with
repo
rted
mem
ory
difficulties;12
mild
[LOC
≤30min,or
GCS13
-15],12
mod
erate[LOC30
min
-24hr
orGCS9-12
],24
severe
[LOC
>24
hror
GCS<9],2n/a.
Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(FreeSurfer).
-Volum
eredu
ctionin
the
hipp
ocam
pus,lateralprefrontal
cortex,thalam
us,andseveral
subregions
ofthecing
ulatecortex
predictedrehabilitationou
tcom
e.
Average
11.5
years.
Con
trols:Non
e.
Warner
2010
aAcute.
3T
Patients:25
patientswith
diffuse
traumatic
axon
alinjury
(18M,
7F;meanage26
.8),[G
CS3-15
].
Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(FreeSurfer).
-Patientsshow
edmajor
glob
albrainatroph
ywith
ameanwho
le-
brainparenchy
mal
volumeloss
of4.5%.
Initial
median1daypo
st-
injury.Follow-upat
median
7.9mon
thspo
st-injury.
Con
trols:22
Con
trols(14M,8F,
meanage32
.4).
-In
patientssign
ificantly
decreased
volumewas
observed
inthe
amyg
dala,hipp
ocam
pus,thalam
us,
CC,putam
en,precuneus,p
ostcentral
gyrus,paracentrallobu
le,and
parietal
andfron
talcortices.
-Caudateandinferior
tempo
ralcortex
didno
tshow
atroph
yin
patients.
-Lossof
who
le-brain
parenchy
mal
volume,atroph
yof
theinferior
parietal
cortex,pars
orbitalis,
pericalcarinecortex,and
154 Brain Imaging and Behavior (2012) 6:137–192
Tab
le2
(con
tinued)
FirstAutho
rYear
Tim
ePost-Injury
Magnet
Sub
jects
AnalysisMetho
dMainFinding
s
supram
arginalgyruspredicted
long
-term
disability.
Warner
2010
bAcute.
3T
Patients:24
patientswith
diffuse
traumatic
axon
alinjury
(16M,
8F;meanage27
.2)[m
eanGCS
6.4;
rang
eno
trepo
rted].
Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(FreeSurfer).
-Bilateralthalam
usvo
lumes
correlated
with
processing
speed.
Initial≤1weekpo
st-injury.
Follow-upat
4-16
mon
ths
post-injury.
Con
trols:Non
e.
-Amyg
dala,hippo
campal,andarang
eof
cortical
region
volumes
were
correlated
with
learning
andmem
ory
performance.
-Thalamus,superior
fron
tal,superior
parietal,andprecun
euscortices
correlated
with
executivefunctio
n.
-Sub
cortical
andcortical
brain
volumes
wereassociated
with
learning
andmem
oryandprocessing
speed,
butno
texecutivefunctio
n.
Yurgelun-Tod
d20
11Not
Reported.
3T
Patients:15
patientswith
oneor
moreTBI(allM;m
eanage34
.9);
14with
atleaston
emild
TBI;1
with
mod
erate/
severe
TBI.[G
CS
notrepo
rted;(m
ild,mod
erateand
severe
accordingto
The
Ohio
State
University
—TBI
IdentificationMetho
d(O
SU-TBI)].14
subjectswere
norm
alon
aclinical
MRI.
Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(FreeSurfer).
-Patientsexhibitedlower
righ
t,left
andtotalfron
tallobe
volumethan
patients.
Con
trols17
controls(allM;mean
age34
.0).
-Right
andtotalcing
ulum
FAcorrelated
with
suicidal
ideatio
nand
impu
lsivity.
Key:M
male,Ffemale,GCSGlasgow
Com
aScale,C
CCorpusCallosum.A
cutedefinedas
under1month,subacuteisdefinedas
>1month
but<6months,andchronicisdefinedas
>6months
Brain Imaging and Behavior (2012) 6:137–192 155
Strangman et al. 2010; Yount et al. 2002), as well as in-creased lateral ventricles, temporal horns of the lateral ven-tricles, and/or ventricular brain ratio (e.g., Anderson et al.1995; Bigler et al. 1997; Gale et al. 1995; Himanen et al.2005; Wilde et al. 2006; Yount et al. 2002). Reduced vol-ume in subcortical gray matter regions has also beenreported (Gale et al. 2005), as has reduced volume in theputamen, precuneus, post-central gyrus, paracentral lobule,parietal cortex, pericalcarine cortex, and supramarginalgyrus (Warner et al. 2010a).
Taken together, these findings suggest that morphometricbrain abnormalities are observed in patients with TBI, al-though many studies did not separate mTBI from moderateand severe TBI. Moreover, in addition to combining mildTBI with moderate and severe cases, the differences amongthe studies reviewed make the interpretations of findingsdifficult, and have led to a sponsored work group meetingin 2009, entitled “the Common Data Elements Neuroimag-ing Working Group.” This work group was established tomake recommendations for “common data elements” thatwill likely be useful for characterizing “radiological featuresand definitions,” which are critically needed to characterizeTBI (Duhaime et al. 2010). This work group was sponsoredby multiple national healthcare agencies, including the De-fense Centers of Excellence (DCOE), The National Instituteof Neurological Diseases and Stroke (NINDS), The Nation-al Institute on Disability and Rehabilitation Research(NIDRR), and the Veterans Administration (VA). This workgroup was also charged with making recommendations forradiological image acquisition parameters that should bestandardized in the quest for delineating brain injuries inTBI, particularly given that different imaging acquisitionparameters have been used for different applications, as wellas for different research studies. Further, if radiologicalimaging is to be used as surrogate endpoints for evaluatingtreatment in clinical trials, then some type of standardizationof the image acquisition parameters is an important consid-eration (Duhaime et al. 2010; Haacke et al. 2010).
Haacke et al. (2010) also notes that brain imaging, partic-ularly using more advanced imaging techniques, affords animportant and unique opportunity to visualize and to quantifybrain injuries in TBI, which is particularly useful in what hedescribes as the 90 % of cases that are categorized as mild. Heand his coworkers note that a systematic characterization ofbrain injuries in TBI will likely lead to increased predictivepower in the area of clinical trials and clinical interventions.The new methods that Haacke and coworkers describe (2010)include, DTI, SWI, MRS, SPECT, PET, Magnetoencephalog-raphy, and Transcranial Doppler. Haacke et al. (2010) alsodiscuss the importance of combining techniques in the samesubjects, such as PET and fMRI.
Selecting optimal protocols has been the focus of inves-tigation in other disorders such as Alzheimer’s disease (e.g.,
Leung et al. 2010) and schizophrenia (e.g., Zou et al. 2005).One has to keep in mind, however, that for multi-centerstudies, not all centers have the most up to date, state-of-the-art imaging, and for this reason some compromisesneed to be made to acquire the best imaging data possibleacross centers, with a focus on more state-of-the-art andexperimental protocols being more possible at researchcenters. Nonetheless, the points raised by this workinggroup (Duhaime et al. 2010; Haacke et al. 2010) are im-portant and there is much room for improvement in thekind of imaging data and analyses performed in the inves-tigation of TBI. For mTBI this becomes even more crucialas subtle, small changes are unlikely to be detected usingmore gross radiological measures of brain pathology. Be-low, we review findings from diffusion imaging studies ofmTBI, an important technology for characterizing diffuseaxonal and focal axonal injuries, and which is among themost promising imaging tools for revealing subtle, smallareas of brain injury in mTBI.
Review of DTI findings in mTBI
DTI is a sensitive measure of axonal injury that is particu-larly important for evaluating small and subtle brain alter-ations that are characteristic of most mTBI. DTI will alsolikely become an important diagnostic tool for individualcases of mTBI, particularly where MR and CT are negative.With respect to the latter, DTI can depict multifocal anddiffuse axonal injuries in individual cases of mTBI. Norma-tive atlases of DTI derived measures that depict anatomicalvariation in healthy controls can also be created so thatindividual cases may then be compared with an atlas inorder to discern the pattern of pathology in an individualcase (e.g., Bouix et al. 2011; Pasternak et al. 2010). We willreturn to the use of atlases in the section on future directionsof research, which follows.
Below we review DTI findings in mTBI. Table 3 liststhose studies that focus on mTBI only, or that include othercategories such as moderate and severe TBI, but nonethelessconduct statistical analyses separately for the mTBI group.Table 4, on the other hand, includes those studies that do notseparate findings in mild TBI from moderate and severeTBI, making findings from these studies more similar tomany of the findings reported for morphometry measures inTable 2, where mild, moderate, and severe TBI were oftennot analyzed separately.
Arfanakis and coworkers (2002) were the first to use DTIto investigate diffuse axonal injuries in mTBI. They inves-tigated FA and MD in anterior and posterior corpus cal-losum, external capsule, and anterior and posterior internalcapsule in 5 patients with acute mTBI (within 24 h of injury)and 10 controls. These brain regions were selected as likely
156 Brain Imaging and Behavior (2012) 6:137–192
showing damage based on post-mortem findings. Resultsshowed no mean diffusivity (MD) differences betweenmTBI patients and controls. Group differences were, how-ever, observed for the corpus callosum and the internalcapsule, where fractional anisotropy (FA) was reduced inthe mTBI group compared with controls. Importantly, thelatter findings are consistent with histopathology findings inmTBI patients who died from other causes (e.g., Adams et al.1989; Bigler 2004; Blumbergs et al. 1994; Oppenheimer1968). These investigators concluded that DTI is an importantearly indicator of brain injury in mTBI and has the potentialfor being an important prognostic indicator of later injury.These investigators were also prescient in noting that longitu-dinal studies are needed to understand the dynamic nature ofmTBI and how it may reflect changes in brain alterations overtime. They did not, however, include a follow up scan ofsubjects in their study.
Inglese and coworkers (2005) used DTI measures to inves-tigate both acute (n020mTBI mean of 4 days post-injury) andchronic mTBI patients (n026 mTBI, mean of 5.7 years post-injury) compared with controls (n029). These investigatorsused FA andMDmeasures of diffusion to detect abnormalitiesin the centrum semiovale, the corpus callosum, and the inter-nal capsule. They reported FA reduction in all three structuresin mTBI compared with controls, as well as increased MD inthe splenium of the corpus callosum, which was higher in theacute sample, but lower in the posterior limb of the internalcapsule compared to the chronic sample. These findings aresimilar to the Arfanakis et al. (2002) study, although theseinvestigators also included patients with chronic TBl. Thesefindings also highlight the importance of detecting brain alter-ations that change over the course of brain injury.
Miles et al. (2008) also investigated mTBI patients onaverage 4 days post-injury and again at 6 months follow upon neuropsychological measures (only). They reported in-creased MD and reduced FA in the centrum semiovale, genuand splenium of the corpus callosum, and in the posteriorlimb of the internal capsule in mTBI compared with con-trols. Moreover, 41% of mTBI cases evinced cognitiveimpairments at baseline, and 33% at follow up 6 monthslater. More specifically, while there was no correlation be-tween baseline measures of MD/FA and cognitive measures,baseline MD and FA correlated significantly with cognitivemeasures at follow up, with FA predicting executive func-tion, and a trend for MD to be associated with reaction time.These findings suggest that there may be predictors ofcognitive function from baseline measures of impairmentas measured on DTI, even when there are no correlationsbetween these measures and cognitive measures at baseline.These investigators also did not include follow up DTI scansat 6 months, only cognitive follow up.
Other investigators examining acute and subacute timeperiods report similar findings. For example, one study of
mild and moderate TBI was conducted within 5 to 14 dayspost-injury in patients who had positive findings on clinicalCT. TBI patients showed reduced FA in the genu of thecorpus callosum (mTBI only) and increased radial diffusiv-ity (RD) in the genu of the corpus callosum in both mild andmoderate TBI (Kumar et al. 2009). Matsushita and cow-orkers (2011) investigated both mild and moderate TBIpatients with a median of 3.5 days post-injury. Theyreported decreased FA in the splenium of the corpus cal-losum in mTBI compared with controls. Other parts of thecorpus showed reduced FA in the moderate group includinggenu, stem, and splenium. These findings suggest moreabnormalities in the moderate TBI group.
Bazarian and coworkers (2007) acquired DTI data frommTBI patients within 72 h of brain trauma and they includeda measure of postconcussive symptoms and quality of lifeassessments. They repeated these assessments 1 month later.Decreased trace was reported particularly in the left anteriorinternal capsule, and increased median FA was reported inthe posterior corpus callosum in mTBI compared with con-trols. Of note, trace measures correlated with postconcussivescores at 72 h and at 1 month post-injury and FA valuescorrelated with 72 h postconcussive score as well as with atest of visual motor speed and impulse control. All subjectsevinced normal findings at time of injury on conventionalCT, suggesting that normal findings on conventional CT andMRI are inadequate for characterizing the kind of injuryobserved in mTBI.
The findings of lower trace and increased FA, however,are more perplexing as the expectation is that if there isdamage to the integrity of white matter then FA will bereduced, and MD likely increased, and this is often reported(see Table 3 and 4). An increase in FA and a decrease in MDmay indicate axonal swelling that occurs early in the courseof injury and may correlate with poor clinical outcome.Bazarian and coworkers (2007) have also suggested thispossibility.
Bazarian and coworkers (2011), in another study thatinvolved one athlete with a concussion and an additional8 athletes with multiple (26-399) subconcussive blows tothe head, reported FA and MD changes that were greatest inthe one concussed athlete, intermediate in the subconcussivegroup, and lowest for controls. Of note, they observed thatboth FA and MD changed in both directions, i.e., increasesand decreases were observed. Kou and coworkers (2010)suggest that increased MD and decreased FA may indicatevasogenic edema, which will likely resolve over time,whereas increased FA and reduced MD may indicate cyto-toxic edema, reflected by axonal swelling and more restrict-ed diffusion of water, thereby explaining the increase in FAand decrease in MD. Thus increased FA and associateddecreased MD, early in the course of brain injury, mayindicate a poor prognosis and hence identifying this group
Brain Imaging and Behavior (2012) 6:137–192 157
Tab
le3
DTIstud
iesin
mild
TBI
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Arfanakis
2002
Acute.
1.5T
Patients:5patients(3
M,
2F;m
eanage35
.6)[G
CS
13-15].Con
ventionalCT
norm
al.
AnalysisandMetho
ds:ROI
Analysis.
5ROIstructures
with
selected
volumes
(i.e.,
representativ
eon
ly),in
the
leftandrigh
themisph
eres:
Anteriorandpo
steriorCC,
external
capsule,anterior
andpo
steriorinternal
capsule.
-Patientsbu
tno
tcontrols
show
edFA
differencesfor
ROIs
betweenho
molog
ous
hemisph
eres.
(24hrs).In2patients,baselin
ecomparedwith
DTIat1mon
thfollo
w-up.
Con
trols:10
controls(5
M,
5F;meanage28
.9years).
Dependent
Measures:Trace,FA
andLI.(LatticeIndex).
-Reduced
FAin
patientsin
theinternalcapsuleandCC
comparedto
controls.
-Notracedifferences
repo
rted.
-Atfollo
w-uptwopatients
show
edim
prov
ementin
somebu
tno
tallareas.
Inglese
2005
Acute
andChron
ic.
1.5T
Patients:46
patients(29M,
17F;meanage36
)[G
CS
13-15].Con
ventional
MRIshow
edcontusions
in5patientsand
hematom
asin
3.
AnalysisandMetho
ds:Who
le-
Brain
Histogram
Analysisand
ROIAnalysis.
3ROIstructures
with
selected
voxels(i.e.,
representativ
e,on
ly):the
centrum
semiovale,CC,
andinternal
capsule.
-Nodifferencesbetween
grou
psusinghistog
ram-
derivedmeasures.
In20
ofthepatients,im
aging
was
performed
atameanof
4.05
days
post-injury.
Con
trols:29
age-
andsex-
matched
controls(15M,
14F;meanage35
).Negativefind
ings
onconv
entio
nalMRI.
Dependent
Measures:FA
and
MD.
-Patientsshow
edredu
cedFA
intheCC,internal
capsule,
andcentrum
semiovale,and
increasedMD
intheCC
andinternal
capsule,
comparedto
controls.
Intheremaining
26,im
aging
was
performed
ameanof
5.7yearspo
st-injury.
-MD
insplenium
ofCCwas
high
erin
patientsscanned
earlierbu
tlower
inthe
posteriorlim
bof
the
internal
capsule,compared
topatientsscannedlater.
Bazarian
2007
Acute.
3T
Patients:6subjects[G
CS
13-15].Con
ventionalCT
norm
al.
AnalysisandMetho
ds:2types
ofwho
lebrainanalyses
performed:VBM
andano
vel,
quantileanalysis.
1)Who
lebrainanalysis.
-Who
lebrainanalysis
show
edlower
1stp
ercentile
tracevalues
inpatients
comparedwith
controls.
72hours:post-concussive(PCS)
andneurobehavioraltestin
g.
Con
trols:6age-
andsex-
matched
orthop
edic
controls;Con
ventional
CTno
rmal.Sub
jects
(patientsandcontrols)
includ
ed8M
and4F;
aged
18-31.
ROIan
alysis:Regions
ofinterestwerealso
analyzed
usingaqu
antileapproach.
2)5ROIs
wereselected
and
manually
outlinedin
nativ
espace:
anterior
internal
capsule,po
steriorinternal
capsule,anterior
CC,
posteriorCC,andexternal
capsule.
-Trace
from
who
lebrain
analysiscorrelated
with
PCSscores
at72
hrsandat
1mon
thpo
st-injury.
At1mon
th:qu
ality
oflifeand
PCSassessments(n=11).
DependentMeasures:FA
,Trace.
PCS,neurob
ehavioralbattery,
andqu
ality
oflifeassessments.
-Trace
from
who
lebrain
analysiswas
lower
inthe
leftanterior
internal
capsule,andshow
edhigh
ermedianFA
values
inthe
posteriorCCin
patients
comparedwith
controls.
158 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
-FA
values
correlated
with
72-h
PCSscore,visual
motor
speed,
andim
pulse
control.
-Finding
ssugg
estlower
traceandincreasedFA
may
indicate
axon
alsw
ellin
gearlyin
thecourse
ofaxon
alinjury,which
correlates
poorly
with
clinical
outcom
e.
Kraus
2007
Chron
ic.
3T
Patients:37
TBIpatients;
20mTBI(8
M,12
F;
meanage35
.85)
[LOC
<30
min],17
mod
erateto
severe
TBI(8
M,9F;
meanage34
.88)
[LOC
>30
min
and/or
GCS
<13
].5in
each
TBIgrou
pwith
otherassociated
trauma.
AnalysisandMetho
ds:ROI
analysis.ROIs
draw
non
standardized
space,FA
maps.
13ROIstructures:anterior
andpo
steriorcorona
radiata,cortico-spinal
tracts,cing
ulum
fiber
bund
les,external
capsule,
forcepsminor
andmajor,
genu
,bod
yandsplenium
oftheCC,inferior
fron
to-
occipitalfasciculus,
superior
long
itudinal
fasciculus
andsagittal
stratum.
-FA
decreaseswereseen
inall13
ROIs
formod
erate/
severe
TBI,bu
ton
lyin
the
cortico-spinal
tract,sagittal
stratum,andsuperior
long
itudinalfasciculus
inmTBI.
≥6mon
thsou
tfrom
injury;
average10
7mon
thsforall
TBIsubjects.
Con
trols:18
controls(7
M,
11F;meanage32
.83).
Dependent
Measures:FA
,RD,
AD,w
hite
matterload
defined
astotalnum
berof
region
swith
redu
cedFA
.Neurocogn
itive
measuresrelevant
toattention,
mem
ory,andexecutive
functio
n.
-Patientswith
mod
erate/
severe
TBIcomparedto
controlsshow
edincreased
AD
andRD.
-Patientswith
mTBI
comparedto
controls
show
edincreasesin
AD,
butno
tRD,sugg
estin
gmyelin
damageno
tpresent.
-Whitematterchangesalon
gaspectrum
from
mTBIto
mod
erate/severe
TBIand
axon
albu
tno
tmyelin
damagemay
characterize
mTBI.
Lipton
2008
Chron
ic.
8mon
thsto
3years
Retrospectiv
estud
yof
consecutiveadmission
sof
mTBI.
1.5T
Patients:17
mTBIpatients
(8M,9F;age26
–70
)[G
CS13
-15,
LOC
<20
min].NegativeCT/
MRIfind
ings
attim
eof
initial
injury.Later,at
timeof
scanning
,1mTBI
subjecth
adasm
allareaof
AnalysisandMetho
ds:Who
le-
brainhistog
ram
analysisand
Vox
el-w
iseanalysisof
retrospectivecases.
Who
le-brain
analysis.
-mTBIpatientsshow
eddecreasedFA
andincreased
MD
inCC,subcortical
white
matter,andinternal
capsules,bilaterally.
Brain Imaging and Behavior (2012) 6:137–192 159
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
sign
alintensity
likelydu
eto
gliosis.
Con
trols:10
controlsof
similarageandgend
erdistribu
tion.
Negative
MRIfind
ings
attim
eof
scan.
Dependent
Measures:FA
,MD.
Histogram
parameters,i.e.,
kurtosis,skew
.
-Who
lebraindecrease
inFA
inmTBIby
histog
ram
analysis.
-Reduced
FAandincreased
MD
sugg
eststhat
mTBI
may
beat
oneendof
the
diffuseaxon
alspectrum
andthepatients
complaining
ofsymptom
smanymon
ths,andin
some
casesseveralyears,po
st-
injury
arecharacterizedby
lack
ofwhite
matter
integrity
inbrainregion
skn
ownto
beaffected
inTBI.
Miles
2008
Acute.
1.5T
Patients:17
patients(11M,
6F;meanage33
.44,
rang
e18
-58)
[GCS13
-15,
LOC<20
min].
AnalysisandMetho
ds:ROI
analysis.
StructuralROIusingcircles:
centra
semiovale,thegenu
andthesplenium
oftheCC,
andthepo
steriorlim
bof
theinternal
capsule.
-mTBIpatientshad
increasedMD
andredu
ced
FAcomparedwith
controls.
Average
4days
(range:1–
10).
Con
trols:29
sex-
andage-
matched
controls(15M,
14F;meanage35
,rang
e18
-61).Not
matched
oneducation.
Dependent
Measures:FA
,MD.
Neuropsycho
logicalmeasures
atbaselin
e(≤24
hour
ofim
aging)
and6mon
thfollo
w-
upfor12
/17mTBI.
-OfthemTBIpatients41
%hadcogn
itive
impairments
atbaselin
eand33
%at
follo
w-up.
Follow-upat
6mon
thsfor
neurop
sycholog
ical
measures.
-Atbaselin
eno
depend
ent
measuresweresign
ificantly
correlated.
-IncreasedMD
and
decreasedFA
atbaselin
ecorrelated
with
cogn
itive
measuresat
follo
w-up.
-Finding
ssugg
estshort-term
predictors
ofcogn
itive
functio
nat
6mon
ths.
(Note:
therewas
noDTIat
6mon
thfollo
w-up).
160 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Niogi
2008
aSub
acuteandChron
ic.
≥1month
(range
1–65
months).
≥1persistent
symptom
for
postconcussive
syndrome.
3T
Patients:34
patients(18M,
16F;meanage37
.4,
rang
e16
-61).[G
CS13
-15
].11
mTBI-negativ
eMRIfind
ings;11
mTBI
micro-hemorrhage;
12mTBI-no
n-specific
white
matterhy
perintensitiesor
chronichemorrhagic
contusions.
AnalysisandMetho
ds:ROI
Analysis.ManualROI,
ellip
soid
shapes
foreach
ROI.
ROIstructures:un
cinate
fasciculus,anterior
corona
radiata,anterior
aspect
oftheinferior
long
itudinal
fasciculus,genu
oftheCC,
thecing
ulum
bund
le,and
thesuperior
long
itudinal
fasciculus.
-Reduced
FAin
mTBI
patientsin
anterior
corona
radiata(41%
ofpatients),
uncinate
fasciculus
(29%),
genu
oftheCC(21%),
inferior
long
itudinal
fasciculus
(21%),and
cing
ulum
bund
le(18%).
Con
trols:26
controls
(19M,77F;meanage
28.3,rang
e17
-58).
Dependent
Measures:FA
.Reactiontim
eto
Attention
NetworkTask(Fan
etal.
2005).
-Reactiontim
eon
asimple
cogn
itive
task
correlated
tonu
mberof
damaged
white
matterstructures.
-(N
ote:
10/11mTBIwith
negativ
eMRIfind
ings
show
edFA
redu
ctions
relativ
eto
controls.)
Niogi
(Extension
of20
08apaper)
2008
bSub
acuteandChron
ic.
3T
Patients:43
patients(28M,
15F;meanage32
.4)
[GCS13
-15].Negative
find
ings
forconv
entio
nal
MRIim
ages
of12
/43
patients.
AnalysisandMetho
ds:
ROIstructures:un
cinate
fasciculus,anterior
corona
radiata,anterior
aspect
oftheinferior
long
itudinal
fasciculus,genu
oftheCC,
thecing
ulum
bund
le,and
thesuperior
long
itudinal
fasciculus.
-mTBIandcontrolsshow
edasign
ificantcorrelation
betweenattentionalcontrol,
measuredby
the
AttentionalNetworkTask,
andFA
inleftanterior
corona
radiata,as
wellas
asign
ificantcorrelation
betweenmem
ory
performance,measuredby
theCVLT
,andFA
inthe
uncinate
fasciculus,
bilaterally.
≥1mon
thpo
st-injury(m
ean
16.9
mon
ths,rang
e1–
53mon
ths).
Con
trols:23
controls
(17M,6F;meanage
29.9)Negativefind
ings
forconv
entio
nalMRI.
ROIanalysis.ManualROI,
ellip
soid
shapes
foreach
ROI.
-In
mTBI,FA
was
redu
cedin
both
thesebrainregion
scomparedwith
controls.
≥1persistent
symptom
for
postconcussive
synd
rome.
Dependent
Measures:FA
.AttentionNetworkTask(Fan
etal.20
05),CaliforniaVerbal
LearningTest(CVLT
)to
test
mem
oryperformance.
-Finding
ssugg
estFA
canbe
used
asabiom
arkerfor
neurocog
nitiv
efunctio
nanddy
sfun
ction.
Rutgers
2008
aSub
acuteandChron
ic.
Median5.5mon
ths(range
0.1-
109.3mon
ths).
1.5T
Patients:21
mTBIpatients
(12M,9
F;m
eanage32
±9)
[GCS13
-15].17
/21
patientsnegativ
efind
ings
AnalysisandMetho
ds:ROI,
tractography.
ROIstructures:cerebrallob
arwhite
matter,cing
ulum
,CC,anterior
andpo
sterior
limbof
theinternal
-Onaverage,patients
show
edredu
cedFA
in9
region
s,predom
inantly
incerebral
lobarwhite
matter,
Brain Imaging and Behavior (2012) 6:137–192 161
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
onMRI;others
contusions
and1extra-
axialhematom
a.
capsules,mesenceph
alon
,brainstem
,and
cerebellu
m.
Cerebrallobarwhite
matter
was
subd
ivided
into
centrum
semiovale,fron
tal
lobe,parietallob
e,tempo
ral
lobe,andoccipitallobe.
cing
ulum
,andCC
comparedto
controls(191
region
s).
Con
trols:11
controls(8
M,
3F;meanage37
±9).No
know
nhistoryof
positiv
eMRIfind
ings.
Dependent
Measures:FA
,vo
lume,ADC,nu
mberand
leng
thof
throug
h-passing
fibers
ineach
ROI,
tractography
measure
ofdiscon
tinuity.
-Discontinuity
onfiber
tracking
was
seen
inon
ly19
.3%
offibers.
-Other
region
sprim
arily
invo
lved
supratentorial
projectio
nfiberbu
ndles,
callo
salfibers,andfron
to-
tempo
ro-occipital
associationfiberbu
ndles.
-The
internal
capsules
and
infratentorial
white
matter
wererelativ
elyun
affected.
-There
was
noassociation
betweentim
einterval
ofscanning
andpo
st-injury
measures.
Rutgers
2008
bChron
ic.
1.5T
Patients:39
TBIpatients
(27M,1
2F;m
eanage34
±12
);24
mild
[GCS13
-15
];9,
mod
erate[G
CS9-
12];and6,
severe
TBI
[GCS≤8
].
AnalysisandMetho
ds:ROI
analysisdo
neby
draw
ingROI
manually
onthepartsof
the
CCon
FAmapsof
each
subject.
ROIstructures:CCgenu
,bo
dy,andsplenium
.-Patientswith
mTBI<
3mon
thspo
st-traum
a(n=
12)show
edredu
cedFA
and
increasedADCin
thegenu
oftheCC.
Average
2.8mon
ths.
Con
trols:10
controls(7
M,
3F;meanage37
±9).No
know
nhistoryof
positiv
eMRIfind
ings.
Dependent
Measures:FA
,ADC,
numberof
fibers.
-Patientswith
mTBI
investigated
>3mon
ths
post-traum
a(n=12
)show
edno
differences.
-Moresevere
traumawas
associated
with
genu
and
splenium
FAredu
ction,
increasedADC,andfewer
numbers
offibers.
162 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Huang
2009
“Post-acute”.
1.5T
Patients:10
patientsmild
TBI(m
eanage25
.0±
11.5;meaneducation
12.7±4.7years)[G
CS13
-15
].7of
10negativ
efind
ings
onconv
entio
nal
CT/M
RI.
AnalysisandMetho
ds:
Probabilistic
tractography,
Tract
Based
SpatialStatistics,
MEG
Fronto-occipitalfasciculus,
superior
long
itudinal
fasciculus,inferior
long
itudinalfasciculus,
uncinate
fasciculus,CC,
andcing
ulum
bund
les.
-Resultspresentedas
acase
series.
Con
trols:14
age-matched
healthysubjects(m
ean
age27
.4±15
.2,mean
education12
.9±
3.2years).
Dependent
Measures:MEG
delta
slow
waves,FA
Autom
ated
Who
leBrain
Analysis.
-MEG
slow
waves
originated
from
cortical
gray
areaswith
redu
ced
DTI.
-In
somecases,abno
rmal
MEG
delta
waves
were
observed
insubjects
with
outob
viou
sDTI
abno
rmality,indicatin
gthat
MEG
may
bemore
sensitive
than
DTIin
diagno
sing
mTBI.
Lipton
2009
Acute.
≤2weeks.
3T
Patients:20
mTBIpatients
(9M,11
F;meanage
33.4)[CGS13
-15,
LOC
<20
min,PTA
<24
hr].
AnalysisandMetho
ds:Vox
el-
wiseanalysis.
Autom
ated
analysisof
who
lebrain.
-Patientshadmultip
leclusters
ofredu
cedFA
infron
talwhite
matter,
includ
ingdo
rsolateral
prefrontal
cortex,with
severalclusters
also
show
inghigh
erMD
ascomparedto
controls.
Con
trols:20
matched
controls(9
M,1
1F;mean
age34
.2).
Dependent
Measures:FA
,MD.
-Reduced
dorsolateral
prefrontal
cortex
FAwas
sign
ificantly
correlated
with
worse
executive
functio
ning
inpatients.
Lo
2009
Chron
ic.
≥2years.
1.5T
Patients:10
patients(5
M,
5F;agerang
e20
-51)
[CGS13
-15].For
those
who
hadconv
entio
nal
MR/CT,
thefind
ings
were
negativ
e.Following
research
scan,1subject
hadasm
allfocalarea
show
inglobargliosis.
AnalysisandMetho
ds:ROI.
ROIs
placed
inthegenu
and
splenium
oftheCC,
posteriorlim
bof
the
internal
capsule,andin
the
pontinetegm
entum.
-Com
paredto
controls,
patientshadlower
FAand
high
erADCin
theleftgenu
oftheCC.
Brain Imaging and Behavior (2012) 6:137–192 163
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Con
trols:10
controls(5
M,
5F;meanage44
).Con
trolswerepatients
referred
forMRIdu
eto
headache
andhadno
historyof
head
trauma.
Dependent
Measures:FA
,ADC.
-Patientsshow
edincreased
FAin
thepo
steriorlim
bof
theinternal
capsule
bilaterally
comparedto
controls.
Geary
2010
Chron
ic.
≥6mon
ths.
3T
Patients:40
mTBIpatients
(17M,23
F;meanage
34.53)
[American
Con
gressof
Rehabilitatio
nMedicine,
1993
criteria].14
subjects
hadprevious
head
trauma.
NegativeMRI/CT
find
ings.
AnalysisandMetho
ds:ROI.
ROIs
structures:anterior
and
posteriorcorona
radiata,
corticospinaltracts
(including
partsof
the
corticop
ontin
etractand
superior
thalam
icradiation),external
capsule,
cing
ulum
,forcepsminor,
forcepsmajor,inferior
fron
to-occipitalfasciculus,
superior
long
itudinal
fasciculus,un
cinate
fasciculus,sagittalstratum,
andbo
dy,genu
,and
splenium
ofCC.
-TBIpatientshaddecreased
FAin
thesuperior
long
itudinalfasciculus,
sagittalstratum,and
uncinate
fasciculus
comparedto
controls.
Con
trols:35
controls
(16M,19
F;meanage
32.54)
matched
topatients
onage,education,
years
ofem
ploy
ment,and
estim
ated
prem
orbid
intelligence.
Dependent
Measures:FA
.-Poo
rperformance
ona
verbal
mem
orytask
was
correlated
with
redu
cedFA
intheun
cinate
fasciculus
andin
thesuperior
long
itudinalfasciculus
inpatients.
Holli(b)
2010
Acute.
≤3weeks.
1.5T
Patients:42
patients(17M,
25F;meanage,38
.8)
[GCS13
-15].DTI
analyses
wereperformed
on34
patients.Negative
CT/M
RIfind
ings.
AnalysisandMetho
ds:Texture
analysis(M
aZda).
Regions
correspo
ndingto
the
mesenceph
alon
,centrum
semiovale,andCC.
-Significant
asym
metries
intextureparameterswere
seen
inallthreeregion
sstud
iedin
thepatients,
theseasym
metries
were
less
prevalentin
the
controls.
Dependent
Measures:FA
,ADC.
-Several
textureparameters
correlated
with
FAand
ADCvalues.
-FA
values
intheleft
mesenceph
alon
were
164 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
10age-
andsex-matched
controls(4
M,6F;mean
age,39
.8).
positiv
elycorrelated
with
verbal
mem
ory.
-ADCin
thebo
dyof
theCC
was
negativ
elycorrelated
with
visual
mem
ory.
Maruta
2010
Sub
acuteto
Chron
ic.
6weeks
and5years(m
ean
2.7years).
3T
Patients:17
patients(10M,
7F;agerang
e20
-52),
[GCS15
with
chronic
postconcussive
synd
rome,
recruitedfrom
clinics].
8no
rmal,9show
edabno
rmalities.
AnalysisandMetho
ds:ROIs
selected
apriori.
Anteriorcorona
radiata,genu
oftheCC,un
cinate
fasciculus,cing
ulum
bund
le,forcepsmajor,and
superior
cerebellar
pedu
ncle.
-Gazeerrorvariability
correlated
with
themean
FAvalues
oftherigh
tanterior
corona
radiata
(ACR)andtheleftsuperior
cerebellarpedu
ncle,and
the
genu
oftheCC.
Con
trols:9controls(6
M,
3F;agerang
e19
-31).No
priorhistoryof
TBI.
Dependent
Measures:FA
.-Reactiontim
ecorrelated
toFA
inthegenu
oftheCC,
andFA
intheleftun
cinate
fasciculus
correlated
with
poorer
verbal
mem
ory.
Mayer
2010
Acute.
≤21
days
(mean12
days).
3T
Patients:22
patients(m
ean
age27
.45)
[GCS13
-15,
LOC<30
min,PTA
<24
hr].
AnalysisandMetho
ds:ROI.
Genu,
splenium
,andbo
dyof
theCC,as
wellas
the
superior
long
itudinal
fasciculus,thecorona
radiata,thesuperior
corona
radiata,theun
cinate
fasciculus,andtheinternal
capsuleforbo
thhemisph
eres.
-TBIpatientshadincreased
FAandredu
cedRD
inthe
CCandseveralleft
hemisph
eretractscompared
tocontrols,i.e.,un
cinate
fasciculus,internal
capsule,
andcorona
radiata.
Con
trols:21
sex-,age-,and
education-matched
controls(m
eanage
26.81).
Dependent
Measures:FA
,AD,
RD.
-Follow-updata
show
edpartialno
rmalizationof
DTIvalues
inseveralwhite
mattertracts.
-The
authorsconcludedthat
initialinjury
toaxon
smight
disrup
tionichemeostasis
andthebalanceof
intra/
extracellu
larwater,which
effectsdiffusionthat
isperpendicularto
theaxon
s.
Zhang
2010
Sub
acute.
30(±2)
days.
3T
Patients:15
stud
ent-athletes
with
mTBI(m
eanage
20.8
years)
[Grade
1MTBIaccordingto
Cantu
AnalysisandMetho
ds:Vox
el-
wisewho
lebrainanalysisand
ROI.
The
CCwas
chosen
asa
prim
aryROIand
subd
ivided
into
thegenu
,bo
dy,andsplenium
.In
-Neither
who
le-brain
analysis
norROIanalysisshow
edsign
ificantalteratio
nin
FA.
Brain Imaging and Behavior (2012) 6:137–192 165
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
2006
.GCSno
trepo
rted].
NegativeMRIfind
ings.
additio
n,therigh
thipp
ocam
pus,leftandrigh
tdo
rsolateral
prefrontal
cortexes
wereevaluated.
Con
trols:15
norm
alstud
ent-athletes
with
nohistoryof
mTBI(m
ean
age21
.3).NegativeMRI
find
ings.The
entire
samplewas
70%M
and
30%F.
Dependent
Measures:FA
,ADC.
-mTBIpatientsshow
edlarger
variability
ofFA
inthegenu
,andbo
dyof
the
CCthan
controls.
-mTBIpatientsshow
eddecreasedADCin
leftand
righ
tdo
rsolateral
prefrontal
cortex
comparedto
controls.
Warner
2010
bAcute.
≤1week
Follow-upat
4-16
mon
ths.
3T
Patients:24
patientswith
diffusetraumatic
axon
alinjury
(16M,8F;mean
age27
.2)[m
eanGCS6.4;
rang
eno
trepo
rted].
AnalysisandMetho
ds:
Autom
ated
Segmentatio
nand
Volum
etricAnalysis
(FreeSurfer).
CC,fornix
body,bilateral
fornix
crus,bilateral
perforantpathway,
cing
ulum
,un
cinate
fasciculus,andinferior
fron
to-occipitalfasciculus.
-The
volumeof
the
hipp
ocam
puscorrelated
with
theFA
inthefornix
body,fornix
crus,perforant
pathway,andcing
ulum
bund
le.
Con
trols:Non
e.Dependent
Measures:Regional
Volum
es,FA
,MD.
-Volum
eof
therigh
tprecun
euscorrelated
with
MD
inthefornix
body
and
righ
tun
cinate
fasciculus.
Note:
Morph
ometricfind
ings
notrelatedto
DTIare
presentedin
Table
2.
Bazarian
2011
Acute.
(72ho
urspo
st-injury).
Prospectiv
ecoho
rtstud
y.
Sub
jectsun
derw
entDTIpre-
andpo
stseason
with
ina3-
mon
thinterval.
(Con
cussed
subjectsun
derw
ent
repeat
testing≤72
hoursof
injury.Sub
jectswho
didno
tsuffer
aconcussion
during
the
season
underw
entrepeat
3T
Patients:One
athlete
concussion
[witn
essed
LOC>20
min,transient
amnesiaor
confusion;
GCSno
trepo
rted]and
8athletes
with
26-399
subcon
cussivehead
blow
s.Con
trols:6controlsub
jects.
2controlsubjectshad
isolated
minor
orthop
edic
injuries.
AnalysisandMetho
ds:Who
lebrainanalysis.FA
andMD
changesweremeasuredin
five
ROIs.
Dependent
Measures:FA
,MD.
Externalcapsule,po
sterior
andanterior
CC,po
sterior
andanterior
limbof
the
internal
capsule.
-FA
changeswerehigh
estfor
theconcussion
subject,
interm
ediary
forsubjects
with
subcon
cussivehead
blow
s,andlowestfor
controls.
-MD
changeswerehigh
est
fortheconcussion
subject,
interm
ediary
forsubjects
with
subcon
cussivehead
blow
sandlowestfor
controls.
166 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
testing≤1weekof
theendof
thesportsseason
3mon
ths
afterinitial
testing).
-The
changesin
FAandMD
intheconcussion
subject
werein
therigh
tcorona
radiataandrigh
tinferior
long
itudinalfasciculus.
-FA
andMD
changesfrom
pre-season
topo
st-season
werein
both
directions,i.e.,
increasedanddecreasedFA
andMD,althou
ghmore
werein
theexpected
direction,
i.e.,decreasedFA
andincreasedMD.
Cub
on20
11Sub
acuteandChron
ic.
≥1mon
th(m
ean115days,SD
104days)forthesports-
relatedconcussion
subjects.
≥1year
forsubjectswith
mod
erateandsevere
TBI.
3T
Patients:10
college
stud
ents
with
concussion
(5M,
5F;meanage19
.7years)
[Diagn
osisbasedon
InternationalCon
sensus
Agreements20
00,GCS
notrepo
rted].2mod
erate
TBIsubjects(m
eanage
20)[G
CS9-12
].3severe
TBI(m
eanage47
.3)
[GCS≤8
].
AnalysisandMetho
ds:TBSS.
Who
lebrainWM
skeleton
(TBSS).
-Con
cussed
subjectsshow
edincreasedMD
inseveral
clusters
intheleft
hemisph
erein
comparison
tocontrols.
Con
trols:10
sex-
andage-
matched
athletes
(5M,
5F;m
eanage20
.4years).
Sex-andage-
matched
controls,forthe2
mod
erate(m
eanage22
)and3severe
TBIsubjects
(meanage46
).
Dependent
Measures:FA
,MD.
-The
largestclusterinclud
edsagittalstratum
(including
theinferior
long
itudinal
fasciculus
andinferior
fron
to-occipitalfasciculus),
retrolenticular
partof
the
internal
capsule,the
posteriorthalam
icradiation
(including
theop
ticradiation),theacou
stic
radiation,
andthesuperior
long
itudinalfasciculus.
-The
second
largestcluster
was
locatedalon
gthe
superior
long
itudinal
fasciculus,anterior
and
superior
tothepartof
the
superior
long
itudinal
fasciculus.
Brain Imaging and Behavior (2012) 6:137–192 167
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Davenpo
rt20
11Chron
ic.
2–5yearsafterblastinjury.
3T
25veterans
with
mTBI
(24M,1F;meanage36
)[LOC<30
min,PTA
<24
hours,and
neurolog
ical
symptom
s;GCSno
trepo
rted).
AnalysisandMetho
ds:20
standard
prob
abilistic
tractography
-based
ROIs.
Forceps
major,forceps
minor,anterior
thalam
icradiations,hipp
ocam
pal
portionof
cing
ulum
,cing
ulum
,corticospinal
tract,inferior
fron
to-
occipitalfasciculus,inferior
long
itudinalfasciculus,
superior
long
itudinal
fasciculus,tem
poralp
ortio
nof
superior
long
itudinal
fasciculus,un
cinate.
-A
glob
alpattern
oflower
white
matterintegrity
was
seen
inblastmTBIpatients
butno
tin
controls.
Con
trols:33
veterans
(28M,5F;meanage
32.5).
Dependent
Measures:FA
.-Patientswith
multip
leblast
injuries
tend
edto
have
agreaterFA
voxelsdecreased
then
individu
alswith
asing
leblastinjury.
Grossman
2011
Chron
ic.
MRIandaneurop
sycholog
ical
battery
either
≤1year
after
injury,or>1year
afterinjury.
3T
Patients:22
patients(14M,
8F;m
eanage38
.2)[G
CS
13-15].
AnalysisandMetho
ds:ROI.
Thalamus
andtheanterior
limb,
genu
,andpo
sterior
limbof
theinternal,the
splenium
oftheCC,and
the
centrum
semiovale.
-Low
ered
DTIandDKI
derivedmeasuresin
the
thalam
usandtheinternal
capsulewereseen
inpatientsexam
ined
oneyear
afterinjury
incomparison
tocontrols.
Con
trols:14
healthy
controls(9
M,5F;mean
age36
.5)matched
topatientsaccordingto
gend
er,age,andform
aleducation
Dependent
Measures:MK,FA
,MD.
-In
additio
nto
theseregion
s,patientsexam
ined
more
than
oneyear
afterinjury
also
show
edsimilar
differencesin
thesplenium
ofthecorpus
callo
sum
and
thecentrum
semiovale.
-Com
bineduseof
DTIand
DKIprov
ides
amore
sensitive
tool
for
identifying
braininjury.
-MK
inthethalam
usmight
beuseful
forearly
predictio
nof
perm
anent
braindamageandcogn
itive
outcom
e.
Henry
2011
Acute.
1–6days
post-con
cussion.
3T
Patients:18
athletes
with
mTBI(allM;meanage
AnalysisandMetho
ds:Vox
el-
basedapproach
(VBA).
Avo
xel-wise2x2repeated-
measuresanalysisof
-mTBIpatientsshow
edincreasedFA
indo
rsal
168 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
For
patients,follo
w-up6mon
ths
later;forcontrols,2n
dscan
18mon
thsafterinitial
scan.
22.08)
[Diagn
osisbased
onAmerican
Academyof
Neurology
Quality
Stand
ards
Sub
committee
andNeurology,19
97;
GCSno
trepo
rted].
variance
(ANOVA)using
SPM8to
thederivedscalar
images
(FA,AD,andMD)
was
used.
corticospinaltractsandin
theCCat
both
timepo
ints
comparedto
control.
Con
trols:10
athletes
(allM;
meanage22
.81).
Dependent
Measures:FA
,MD,
AD
-mTBIpatientsshow
edincreasedAD
intherigh
tcorticospinaltractsat
both
timepo
intscomparedto
controls.
.-mTBIpatientsshow
eddecreasedMD
inthe
corticospinaltractsandCC
atbo
thtim
epo
ints
comparedto
controls.
Lange
2011
Sub
acute.
6-8weeks.
3T
Patients:60
mTBIpatients
(43M,17
F;meanage
30.8)[Closedhead
injury
and;LOC>1minute,PTA
>15
minutes,G
CS<15
,or
abno
rmality
onCT].n=
41negativ
eCTfind
ings,
N=15
positiv
eCT
find
ings,andn=4no
tordered.
N=21
with
postconcussive
disorder
basedon
ICD-10;
N=39
with
outpo
stconcussive
disorder.
AnalysisandMetho
ds:ROIs.
Genu,
body,andsplenium
oftheCC.
-Nodifferenceson
FAor
MDbetweenmTBIpatients
andcontrols.
Con
trols:34
controls
(25M,9F;meanage
37.1)with
orthop
edic/
soft-tissueinjuries.N=0
positiv
efind
ings,n=3
negativ
efind
ings,andn=
31no
tordered.
52.9%%
ofcontrolsmet
ICD-10
criteriaforpo
stconcussive
disorder.
Dependent
Measures:FA
,MD.
-mTBIpatientsrepo
rted
morepo
st-concussion
symptom
sthan
controls.
-mTBIpatientsshow
eda
nonsignificant
trend
increase
inMD
inthe
Brain Imaging and Behavior (2012) 6:137–192 169
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
splenium
oftheCC
comparedto
controls.
MacDon
ald
2011
Sub
acute.
≤90
days.
1.5T
Patients:63
patients(allM,
medianage24
)[D
iagn
osisbasedon
U.S.
military
clinical
criteria
fortraumatic
braininjury;
GCSno
trepo
rted].
AnalysisandMetho
ds:ROI.
Genuandsplenium
ofthe
CC,therigh
tandleft
middlecerebellar
pedu
ncles,therigh
tand
left
cerebral
pedu
ncles,the
righ
tandleftun
cinate
fasciculi,andtherigh
tand
leftcing
ulum
bund
les.
-TBIsubjectsshow
eddecreasedRAin
themiddle
cerebellarpedu
ncles,
cing
ulum
bund
les,andin
therigh
torbitofron
talwhite
matter.
21controls(allM,median
age31
)expo
sedto
blasts,
butno
newith
sustained
TBI.
Dependent
Measures:Relative
anisotropy
(RA),MD,AD,
RD.
-Follow-upDTIscansin
47TBIsubjects6-12
mon
ths
latershow
edpersistent
abno
rmalities.
Matthew
s20
11Self-repo
rted
historyof
concussion
.3T
Patients:22
patientswith
ahistoryof
blast-related
TBI.[LOC<30
min,PTA
<25
hr,CGS13
-15].11
TBIpatientswith
Major
DepressiveDisorder
(MDD)(allM,meanage
26.8).11
TBIpatients
with
outMDD
(allM,
meanage30
.3).
AnalysisandMetho
ds:VBA
-Patientswith
MDD
show
edlower
FAin
corona
radiata,
CCandsuperior
long
itudinalfasciculus
comparedto
non-MDD
patients.
Con
trols:Non
e.-FA
inthesuperior
long
itudinalfasciculus
correlated
negativ
elywith
depressive
symptom
s.
Matsushita
2011
Acute.
Median3.5days.
1.5T
Patients:20
patients(18M,
2F;meanage39
.3):9
mild
TBI[G
CS13
-15,
LOC<30
min,PTA
<24
hr].11
mod
erateTBI
[LOC≥3
0min
and/or
GCS9-12
].
AnalysisandMetho
ds:ROI.
Genu,
stem
,andsplenium
oftheCCandthecorona
radiata,anterior
limbof
the
internal
capsule,po
sterior
limbof
theinternalcapsule,
fron
talwhite
matter,and
occipitalw
hitematterof
the
periventricularwhite
matter.
-Decreased
FAin
the
splenium
oftheCCin
mTBIcomparedto
controls.
Con
trols:27
matched
controls(13M,14
F;
meanage42
.9).
Dependent
Measures:FA
.-Decreased
FAin
thegenu
,stem
andsplenium
ofthe
CCin
mod
erateTBI
comparedto
controls.
170 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Messe
2011
Acute.
7-28
days
(sub
acuteph
ase),
Sub
acute.
3-4mon
ths(lateph
ase).
1.5T
Patients:23
patients[LOC
<30
min,CGS13
-15,
PTA
<24
hr,or
neurolog
ical
symptom
s];
11with
good
outcom
e(8
M,3
F;m
eanage27
.8),
12with
poor
outcom
e(7
M,5
F;m
eanage31
.3).
7of
11with
good
outcom
ehadnegativ
eMRIfind
ings
and6ou
tof
12po
orou
tcom
ehad
negativ
eMRIfind
ings.
AnalysisandMetho
ds:
VBM,TBSS.
Who
leBrain.
-MD
was
high
erin
poor
outcom
epatientscompared
with
bothcontrolsandgo
odou
tcom
epatients,in
the
forcepsmajor
andminor
oftheCC,inferior
fron
to-
occipitalfasciculus,and
inferior
long
itudinal
fasciculus.
Con
trols:23
controls
(12M,11
F;meanage
30).
Dependent
Measures:FA
,MD,
AD,RD.
-MD
was
high
erin
poor
outcom
epatientscompared
tocontrolsin
superior
long
itudinalfasciculus
and
corticospinaltract,andleft
anterior
thalam
icradiation.
-Nodifference
indiffusion
variableswas
foun
dbetweenpatientswith
good
outcom
eandcontrols.
Smits
2011
Sub
acute.
Average
30.6
days.
3T
Patients:19
patients
includ
ed(10M,9
F;m
ean
age26
)[G
CS13
-15].
AnalysisandMetho
ds:TBSS.
Who
leBrain.
-There
wereno
differences
inMD
betweenpatients
andcontrols.
Con
trols:12
controls(8
M,
4F;meanage28
).Dependent
Measures:FA
,MD.
-Decreased
FAwas
inthe
righ
ttempo
ralsubcortical
white
matterin
patients
comparedto
controls.
These
region
sinclud
edinferior
fron
to-occipital
fasciculus.
-IncreasedMD
was
correlated
with
severity
ofpo
st-con
cussivesymptom
sin
theinferior
fron
to-
occipitalfasciculus,inferior
long
itudinalfasciculus
and
thesuperior
long
itudinal
fasciculus.
Brain Imaging and Behavior (2012) 6:137–192 171
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
-Decreased
FAcorrelated
with
severity
ofpo
st-
concussive
symptom
sin
theun
cinate
fasciculus,
inferior
fron
to-occipital
fasciculus,internal
capsule,
CC,parietal
subcortical
white
matterandfron
tal
subcortical
white
matter.
-Microstructural
changesin
white
matterarelik
elya
neurop
atho
logicalsubstrate
ofpo
stconcussive
synd
rome.
Spo
nheim
2011
Chron
ic.
Several
Mon
ths.
1.5T
Patients:9patients(allM;
meanage33
.7)
[Diagn
osisbasedon
American
Con
gressof
Rehabilitatio
nMedicine
Special
InterestGroup
onMild
Traum
atic
Brain
Injury
andtheconcussion
gradingsystem
bythe
American
Academyof
Neurology
;GCSno
trepo
rted].
AnalysisandMetho
ds:ROI.
Forceps
major,forcepsmajor
andanterior
thalam
icradiations.
-EEG
phasesynchron
ywas
less
inlateralfron
tal
region
sin
patients.
Con
trols:8controls(allM;
meanage30
.3).
Dependent
Measures:
Electroenceph
alog
ram
(EEG)
phasesynchron
izationandFA
.
-EEG
phasesynchron
ywas
correlated
with
FAin
the
leftanterior
thalam
icradiations
andtheforceps
minor
inpatientsbu
tno
tcontrols.
McA
llister
2012
Acute.
Preseason
and≤10
days
ofa
diagno
sedconcussion
.
3T
Sub
jects:10
athletes
(allM;
agerang
e15
-23)
with
mTBI[D
iagn
osisby
certifiedathletic
traineror
team
physician;
GCSno
trepo
rted].
AnalysisandMetho
ds:Strain
andstrain
rate
calculations
basedon
Dartm
outh
Sub
ject-
SpecificFEHeadmod
el(helmetswornto
record
head
impactsdu
ring
play
onthe
field).DTImapsforCC.
CC.
-Chang
ein
FAcorrelated
with
meanandmaxim
umstrain
rate.
172 Brain Imaging and Behavior (2012) 6:137–192
Tab
le3
(con
tinued)
FirstAutho
rYear
Typ
eof
Study
(tim
epo
st-injury)
Magnet
Sub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Non
e.Dependent
Measures:FA
,MD,
Strain,
StrainRate.
-Chang
ein
MD
correlated
with
meanandmaxim
umstrain.
-Chang
ein
MD
correlated
with
leng
thof
injury-to-
imaginginterval,bu
tchange
inFA
didno
t.
Key:M
Male,
FFem
ale,
PTA
Posttraumatic
Amnesia,
LOC
Lossof
Con
sciousness,GCSGlascow
Com
aScale,CC
Corpu
sCallosum,ADC
ApparentDiffusion
Coefficient,FA
Fractional
Anisotrop
y,MDMeanDiffusivity,R
DRadialD
iffusivity,A
DAxialDiffusivity,D
KIDiffusionalKurtosisIm
aging,DTIDiffusion
TensorIm
aging,MKMeanKurtosis,ROIRegionof
Interest,T
BSS
Tract-Based
SpatialStatistics
Thistablepresentsfind
ings
derivedfrom
diffusiontensor
imaging(D
TI)includ
ing:
Trace,M
eanDiffusivity
(MD),FractionalA
nisotrop
y(FA),AxialDiffusivity
(AD),RadialD
iffusivity
(RD),and
MeanKurtosis(M
K).In
stud
ieswhere
non-DTIdiffusionweigh
tedim
aging(D
WI)was
also
presentedweinclud
ethemainmeasure,A
pparentDiffusion
Coefficient
(ADC).In
thedescriptionof
subjects,w
eprov
ide,whenavailable,theGlasgow
Com
aScale(G
CS)score,Lossof
Con
sciousness
(LOC)du
ratio
n,andPosttraumaticAmnesia(PTA
)du
ratio
n.Acuteisdefinedas
under1mon
th,
subacute
isdefinedas
>1mon
thbu
tless
than
6mon
ths,andchronicisdefinedas
>6mon
ths
Brain Imaging and Behavior (2012) 6:137–192 173
Tab
le4
DTIstud
ieswith
differentDTIseveritiesinclud
ingmild
FirstAuthor
Year
Typ
eof
Study
[tim
epo
st-injury]
MagnetSub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
Huism
an20
04Acute
<7days.
1.5T
Patients:20
patients(15M,5
F;
meanage31
).[G
CS4-15
].AnalysisandMetho
ds:ROI
Analysis.
Posterior
limbof
theinternal
capsule(bilaterally
)andthe
splenium
oftheCC.
-FA
values
weresign
ificantly
redu
cedin
theinternal
capsule.
Con
trols:15
healthysubjects
with
amatchingage
distribu
tion(m
eanage35
).
Dependent
Measures:FA
,ADC.Measurementsin
thethalam
usandpu
tamen
served
asinternal
references.
-FA
values
weresign
ificantly
correlated
with
GCSandRankin
scores
fortheinternal
capsule
andsplenium
.
-ADCvalues
weresign
ificantly
redu
cedwith
inthesplenium
.
-Correlatio
nbetweenFA
and
clinical
markers
was
betterthan
that
fortheADCvalues.
-ADCof
theinternal
capsulewas
notcorrelated
with
theGCSand
Rankinscores.
Salmon
d20
06Chron
ic>6mon
ths.
3T
Patients:16
patients(13M,3
F;
meanage32
)[M
edianGCS7,
rang
e3-13
].
AnalysisandMetho
ds:Vox
elbasedanalysis.
Mapsof
FAandMD
were
calculated
from
the
diffusion-weigh
tedim
ages.
-Major
white
mattertractsand
associationfibersinthetempo
ral,
fron
tal,parietal
andoccipital
lobesshow
edbilateraldecreases
inFA
.
Con
trols:16
controls[12M,
3F;meanage34
].Dependent
Measures:FA
,MD.
-Increasesin
MD
werefoun
din
widespreadareasof
thecortex.
-A
positiv
ecorrelationwas
foun
dbetweenMD
andim
pairmentof
learning
andmem
oryin
theleft
posteriorcing
ulate,left
hipp
ocam
palform
ationandleft
tempo
ral,fron
talandoccipital
cortex.
Benson
2007
Acute
toChron
ic.
Variablefrom
days
toyears
(mean35.3months,range
3days
to15
years).
1.5T
Patients:
AnalysisandMetho
ds:
Who
le-brain
analysisusing
histog
ram
shapeparameters.
-White
matterwho
le-brain
FAhistog
ramsshow
edadecrease
inFA
inTBIpatientswith
noov
erlapbetweenpatient
and
controls.
20patients(13M,7
F;m
eanage
35.5)[G
CS3-15
].6mTBIand
14mod
erate/severe
TBIAll
CT+except
N=3mTBI.
White
matterwho
le-brain
FAhistog
ram
parameters.
-Histogram
shapeparameters
correlated
with
each
otherand
with
admission
GCSandPTA
.
Con
trols:14
age-matched
controls(m
eanage27
.5).
Dependent
Measures:FA
,Kurtosis,Skewness.
-MeanFA
was
thebestpredictorof
PTA
duratio
n(r=-.64
)andGCS
(-.47).
174 Brain Imaging and Behavior (2012) 6:137–192
Tab
le4
(con
tinued)
FirstAuthor
Year
Typ
eof
Study
[tim
epo
st-injury]
MagnetSub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
-40
%%
ofPTA
variance
inADC
was
accoun
tedforby
FA.
-Mostof
theFA
decrease
was
likelyaccoun
tedforby
increased
diffusionin
theshortaxis
dimension
,reflectin
gdy
smyelin
ationandaxon
alsw
ellin
g.
-The
correlationof
injury
severity
with
changesin
FAsugg
estsa
role
forDTIin
earlydiagno
sis
andprog
nosisof
TBI.
Kum
ar20
09Acute.
5–14
days.
1.5T
Patients:83
patients[62M,
21F;meanage34
.25].26
mTBI[G
CS13
–15
,mean
14.5].57
mod
erateTBI[G
CS
9–12
].Allwerepo
sitiv
efor
clinical
find
ings
onCTat
injury.
AnalysisandMetho
ds:ROI.
CC(divided
into
7segm
ents).
-Decreased
FAin
genu
inmTBI
comparedto
mod
erateTBI,and
inbo
thTBIgrou
pscompared
with
controls.
Con
trols:33
age-
andsex-
matched
controls(22M,1
1F;
meanage32
.1).
Dependent
Measures:FA
,RD,
MD,AD.
-Decreased
FAin
splenium
inmod
erateTBIcomparedto
controls.
-Decreased
MD
inmod
erateTBI
comparedto
controls.
-IncreasedRD
ingenu
and
splenium
inmTBIandmod
erate
TBIcomparedto
controls.
-IncreasedRD
inmidbo
dyCCin
mod
erateTBIcomparedto
controls.
-Decreased
AD
ingenu
inmod
erateTBIcomparedto
controls.
-Poo
rneurop
sycholog
ical
outcom
eat
6-mon
thswas
associated
with
DTabno
rmalities
intheCC.
Levin
2010
Chron
ic.
Post-injury
interval
871.5days.
3T
Patients:37
veterans
(meanage
31.5)with
mild
ormod
erate
TBI[G
CSno
tprov
ided].5
AnalysisandMetho
ds:
Tractog
raph
y,standard
sing
le-
sliceROImeasurement,and
Non
-tractog
raph
yROI
protocolsinclud
edthetotal
CC(including
genu
,bo
dy,
-Nogrou
pdifferencesin
FAand
ADCwerefoun
d.How
ever,FA
oftheleftandrigh
tpo
sterior
Brain Imaging and Behavior (2012) 6:137–192 175
Tab
le4
(con
tinued)
FirstAuthor
Year
Typ
eof
Study
[tim
epo
st-injury]
MagnetSub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
hadpo
sitiv
efind
ings
onclinical
MRI.
voxel-basedanalysis.
Manually
-tracedROIs
ona
slice-by
-slicebasis,and
quantitativefibertractography
toassess
whitematterintegrity
splenium
,and
total),and
the
righ
tandleftanterior
and
posteriorlim
bsof
the
internal
capsule.
internal
capsuleandleft
corticospinaltractpo
sitiv
ely
correlated
with
totalwords
consistently
recalled.
Con
trols:15
veterans
with
outa
historyof
TBIor
expo
sure
toblast(m
eanage31
.4),
includ
ing7subjectswith
extracranialinjury
(meanpo
st-
injury
interval
919.5days),
and8who
wereun
injured.
Dependent
Measures:FA
,ADC
-ADCfortheleftandrigh
tun
cinate
fasciculiandleft
posteriorinternal
capsulewas
negativ
elycorrelated
with
this
measure
ofverbal
mem
ory.
-Correlatio
nsof
DTIvariables
with
symptom
measureswere
non-
sign
ificantandinconsistent.
Hartik
ainen
2010
Acute.
3weeks.
1.5T
Patients:18
patients(6
M,1
2F;
11asym
ptom
atic
and7
symptom
atic;meanage43
)with
mild
andmod
erateTBI.
Allmod
eratesubjectshad
positiv
efind
ings
onclinical
CTor
MRI.
AnalysisandMetho
ds:ROI
Thalamus,internal
capsule,
centrum
semiovale,and
mesenceph
alon
.
-Sym
ptom
atic
patientshadhigh
erFA
andlower
ADCin
the
mesenceph
alon
then
asym
ptom
atic
patients.
Con
trols:Non
e.Dependent
Measures:FA
,ADC
-Non
eof
theotherFA
orADC
measurementsweresign
ificantly
differentbetweensymptom
atic
andasym
ptom
atic
patients.
Little
2010
Chron
ic.
≥12mon
ths.
3T
Patients:12
patientswith
mTBI
(meanage31
.2).12
mod
erate
tosevere
TBI(m
eanage33
.3)
Severity
was
basedon
duratio
nof
Lossof
Con
sciousness
andPost-
Traum
atic
Amnesia.4mTBI
and9mod
erateto
severe
patientshadpo
sitiv
efind
ings
onMRIor
CT.
AnalysisandMetho
ds:ROI.
12ROIs
incortical
andCC
structures
and7subcortical
ROIs
(anterior,ventral
anterior,ventrallateral,
dorsom
edial,ventral
posteriorlateral,ventral
posteriormedial,and
pulvinar
thalam
icnu
clei).
-TBIpatientsshow
edredu
ctions
inFA
intheanterior
andpo
sterior
corona
radiata,forcepsmajor,the
body
ofCC,and
fibersidentified
from
seed
voxelsin
theanterior
andventralanterior
thalam
icnu
clei.
Con
trols:12
age-andeducation-
matched
controls(m
eanage
30.8).
Dependent
Measures:FA
.-FA
from
cortico-corticoandCC
ROIs
didno
taccoun
tfor
sign
ificantvariance
inneurop
sycholog
ical
functio
n.How
ever,FA
from
thethalam
icseed
voxelsdidaccoun
tfor
176 Brain Imaging and Behavior (2012) 6:137–192
Tab
le4
(con
tinued)
FirstAuthor
Year
Typ
eof
Study
[tim
epo
st-injury]
MagnetSub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
variance
inexecutivefunctio
n,attention,
andmem
ory.
Singh
2010
Acute
toChron
ic.
Mean~1
mon
th.
1.5T
Patients:12
patients(m
eanage
28)with
mild
tomod
erate
braininjuries
(lossof
consciou
snessfor<1h)
prim
arily
dueto
falls,assault
ortrafficaccidents.
AnalysisandMetho
ds:ROI
analysis.
Differences
intheFA
maps
betweeneach
TBIsubject
andthecontrolgrou
pwere
compu
tedin
acommon
spaceusingat-test,
transformed
back
tothe
individu
alTBIsubject's
head
space,andthen
thresholdedto
create
ROIs
that
wereused
tosorttracts
from
thecontrolgrou
pand
theindividu
alTBIsubject.
Tractcoun
tsforagivenROI
ineach
TBIsubjectwere
comparedto
grou
pmeanfor
thesameROIin
orderto
quantifytheim
pactof
injury
alon
gaffected
pathways.
The
sameprocedurewas
used
tocompare
theTBI
grou
pto
thecontrolg
roup
inacommon
space.
-With
inTBIsubjectsaffected
pathwaysinclud
edhipp
ocam
pal/
fornix,inferior
fron
to-occipital,
inferior
long
itudinalfasciculus,
CC(genuandsplenium
),cortico-
spinal
tractsandtheun
cinate
fasciculus.
Con
trols:10
age-matched
volunteers
(meanage27
).No
evidence
ofneurolog
ical
diseaseon
MRim
aging.
Dependent
Measures:FA
,MD,
AD,RD.
-Mostof
theseregion
swerealso
detected
inthegrou
pstud
y.
Brand
stack
2011
Acute.
Mean7days
1.5T
Patients:22
patients(G
CS6-15
)admitted
totheem
ergency
departmentforTBI.On
clinical
MRI7werenegativ
eforfind
ings,15
show
edcortical
contusions
ortraumatic
axon
alinjury.
AnalysisandMetho
ds:ROI
analysis.
Measurementswere
performed
at46
different
locatio
nsinclud
ingbilateral
cortical,subcortical
and
deep
brainstructures.
-TBIpatientswith
andwith
out
visiblelesion
sdidno
tshow
any
differencesin
ADCvalues
but
both
grou
psdiffered
from
the
controlsin
severalcortical
and
deep
brainregion
s.
Con
trols:14
healthysubjects.
Dependent
Measures:ADC.
-IncreasedADCvalues
were
common
inTBIgrou
psbu
tdecreasedADCvalues
were
relativ
elyun
common
.
-RegionalADCvalues
andthe
numberof
abno
rmal
region
sdid
notcorrelatewith
either
GCSor
Brain Imaging and Behavior (2012) 6:137–192 177
Tab
le4
(con
tinued)
FirstAuthor
Year
Typ
eof
Study
[tim
epo
st-injury]
MagnetSub
jects(N
,gend
er,age)
DTIAnalysisMetho
dand
Dependent
Measures
Brain
Region(s)
MainFinding
(s)
GOS(G
lasgow
outcom
escale)-E
scores.
Ljung
qvist
2011
Acute.
With
in11
days
andat
6mon
thspo
st-injury.
1.5T
Patients:8patientswith
TBI
(4M,4F;meanage36
.4)
[GCS3-14
].Allpatientswere
CTnegativ
e.
AnalysisandMetho
ds:ROI
analysis.
Polyg
onal
ROIs
were
manually
placed
intheCC.
-Patientsshow
edaredu
ctioninFA
inthecorpus
callo
sum
inthe
acuteph
asecomparedto
controls.
Con
trols:6controls(3
M,3F;
meanage33
.1).
DependentMeasures:FA
,Trace,
AD,RD.
-There
was
nosign
ificantchange
inAD,RD,or
Trace.
-At6mon
ths,asign
ificant
redu
ctionin
FAandasign
ificant
increase
inTrace
andAD
inpatientscomparedto
controls.
Yurgelun-
Tod
d20
11Not
repo
rted.
3T
Patients:15
Mveterans
with
TBI(m
ild,mod
erateand
severe
accordingto
The
Ohio
State
University
-TBI
IdentificationMetho
d).
AnalysisandMetho
ds:ROI
analysis.
ROIs
formajor
white
matter
tractswerecreatedforthe
genu
andthecing
ulum
using
theJohn
sHop
kins
University
White
Matter
LabelsAtlas.
-TBIpatientsshow
edadecrease
inFA
values
intheleftcing
ulum
,andleftandtotalgenu
compared
tocontrols.
Con
trols:17
Mcontrols:10
civilians,6veterans.
Dependent
Measures:FA
,MD.
-The
TBIgrou
pshow
edgreater
impu
lsivity
-Total
andrigh
tcing
ulum
FApo
sitiv
elycorrelated
with
current
suicidalideatio
nandmeasuresof
impu
lsivity.
Key:M
Male;FFem
ale;GCSGlasgow
Com
aScale,C
CCorpu
sCallosum,A
DCApp
arentD
iffusion
Coefficient,FAFractionalA
nisotrop
y,MDMeanDiffusivity,R
DRadialD
iffusivity,A
DAxial
Diffusivity,ROIRegionof
Interest
Thistablepresentsfind
ings
derivedfrom
Diffusion
TensorIm
aging(D
TI),including
:Trace,M
eanDiffusivity
(MD),FractionalAnisotrop
y(FA),AxialDiffusivity
(AD),RadialDiffusivity
(RD),
andMeanKurtosis(M
K).In
stud
ieswhere
non-DTIdiffusionweigh
tedim
aging(D
WI)was
also
presented,
weinclud
ethemainmeasure,A
pparentD
iffusion
Coefficient
(ADC).In
thedescription
ofthesubjects,weprov
ide,whenavailable,theGlascow
Com
aScale(G
CS)score,Lossof
Con
sciousness
(LOC)du
ratio
n,andPosttraumatic
Amnesia(PTA
)du
ratio
n.Acuteisdefinedas
under
1mon
th,subacute
isdefinedas
>1mon
thbu
tgreaterthan
6mon
ths,andchronicisdefinedas
>6mon
ths.
178 Brain Imaging and Behavior (2012) 6:137–192
early post-injury may make early interventions possible.Here, measures such as free water, which characterize wateras belonging to tissue versus extracellular water (e.g.,Pasternak et al. 2009), may be helpful in determining intra-cellular versus extracellular edema, i.e., cytotoxic versusvasogenic edema. Such measures are just beginning to beused in mTBI by our group (e.g., Pasternak, Kubicki and etal. 2011a; b).
A further issue is whether or not an observed increase inFA early in the course of injury is predictive of pooreroutcome, or if an increase in FA at any time over the courseof the injury is a poor indicator of outcome. This question isrelevant, particularly given that some investigators do notreport increased FA at 24 h post-injury (e.g., Arfanakis et al.2002), while others report increased FA at 72 h (Bazarian etal. 2007). Of interest here, Henry et al. (2011) scanned 18athletes with mTBI at 1 to 6 days post-injury, later in post-injury than either the Arfanakis or Bazarian studies, andthey also reported increased FA and decreased MD in dorsalcorticospinal tracts and in the corpus callosum at both base-line and at 6 month follow up. Further, Mayer et al. (2010)scanned mTBI patients 21 days post-injury, on average12 days post-injury. They, too, reported increased FA andreduced radial diffusivity (RD) in the corpus callosum inmTBI patients compared with controls. Finally, a study byHartikainen et al. (2010) compared 11 asymptomatic TBIpatients with 7 symptomatic TBI patients, which includedboth mild and moderate cases, with moderate TBI patientsall having abnormal findings on clinical CT or MRI. Incomparing these two groups, the symptomatic patientsshowed increased FA and lower ADC in the mesencephaloncompared with the asymptomatic patients. Thalamus, inter-nal capsule, and centrum semiovale did not differentiatebetween these two groups. There was no control group.Nonetheless, the notion that more symptomatic patientsshowed increased FA is important and could serve as abiomarker for predicting those patients who have, or willhave, a poorer outcome.
The implications of increased FA and reduced MD, andwhether these measures are indicators of poor outcomeclearly needs further investigation. A longitudinal studywould help to follow the natural progression of brain injuryin mTBI over time in order to determine the significance ofincreased versus decreased FA that is observed in mTBI.Work by Lipton and coworkers (see empirical study in thisspecial issue) have investigated, in a longitudinal design,areas of increased and decreased FA, sometimes observed inthe same patient, and they suggest that increased FA may beassociated with a compensatory mechanism or neuroplastic-ity in response to brain injury, but it is not necessarily adirect reflection of brain injury per se. Further work isneeded to determine the biological meaning and implica-tions of increased and decreased FA, as the story may be
more complex than interpretations given thus far to explainthese phenomenon.
Other noteworthy findings of brain injury in mTBI in-clude reduced FA in frontal white matter, including dorso-lateral prefrontal cortex, where Lipton et al. (2009) reportedseveral clusters of increased MD, which were correlatedwith worse executive functioning in a group of acute mTBIpatients compared with controls. Other findings includeabnormalities in mesencephalon, centrum semiovale, andcorpus callosum (Holli et al. 2010b), with findings for themesencephalon being correlated with verbal memory inmTBI patients. Additionally, reduced FA has been reportedin the anterior corona radiata, in the genu of the corpuscallosum, and in the left superior cerebellar peduncle(Maruta et al. 2010). In the latter study, gaze positionalerrors in an eye movement task were correlated with meanFA values of the right anterior corona radiata, the left supe-rior cerebellar peduncle, and the genu of the corpus cal-losum, all tracts that support spatial processing and areinvolved in attention, leading these investigators to suggestthat gaze errors might be a useful screening tool for mTBI.
Kraus and coworkers (2007) focused on chronic patientswho had mild (n020), moderate and severe (n017) injury,and 18 controls. They investigated FA as well as radialdiffusivity (RD), axial diffusivity (AD), and white matterload; the latter defined as the total number of regions withreduced FA. She and her coworkers reported FA decreasesin all thirteen brain regions examined in the moderate tosevere group. In the mTBI group, FA was reduced in onlythe corticospinal tract, sagittal striatum, and superior longi-tudinal fasciculus. Both AD and RD were increased inseveral white matter regions in the moderate to severegroup, whereas in the mTBI group only increases in ADwere observed, suggesting that myelin damage is not presentin mTBI but is present in moderate to severe TBI. Theseinvestigators concluded that white matter changes that indi-cate diffuse axonal injury in TBI show alterations along aspectrum from mild to severe TBI.
Lipton et al. (2008) also investigated more chronicpatients. They compared 17 mTBI patients who had nega-tive CT/MRI findings at the time of injury, with 10 controlswho also showed negative findings on MRI. One of themTBI subjects subsequently was observed to develop asmall area of increased signal intensity, likely the result ofgliosis. Decreased FA and increased MD were observed inthe corpus callosum, subcortical white matter, and in theinternal capsules, bilaterally. These brain regions are similarto those observed as abnormal in the acute studies notedabove (i.e., Arfanakis et al. 2002; Inglese et al. 2005;Bazarian et al. 2007; Miles et al. 2008).
Niogi and colleagues (2008a) investigated a group ofpatients 1 to 65 months following injury; all characterizedas having more than one symptom of postconcussive
Brain Imaging and Behavior (2012) 6:137–192 179
syndrome. Subjects included 34 mTBI patients that wereseparated into those with negative MRI findings (n011),those with micro-hemorrhage (n011), and those with whitematter hyperintensities or chronic hemorrhagic contusions,compared with controls (n026). They investigated sixregions of interest and reported reduced FA in the anteriorcorona radiata (41 %), the uncinate fasciculus (29 %), thegenu of the corpus callosum (21 %), the inferior longitudinalfasciculus (21 %), and the cingulum bundle (18 %). Reac-tion time was correlated with a number of damaged whitematter structures and it was noted that 10 of the 11 of themTBI patients with negative MRI findings showed reduc-tions in FA relative to controls. In a slightly larger sample ofmTBI patients, this group also reported significant correla-tions between attentional control and FA reduction in the leftanterior corona radiata, as well as significant correlationsbetween memory performance and reduced FA in the unci-nate fasciculus, bilaterally (Niogi et al. 2008b). These cog-nitive correlates with white matter fiber tract abnormalitiesfindings are consistent with what would be expected giventhe function of these tracts, further suggesting that FA maybe useful as a biomarker for neurocognitive function anddysfunction in mTBI.
Given the different magnet strengths, with some con-ducted on a 1.5 T magnet, and others conducted on a 3 Tmagnet (see Tables 3 and 4), as well as differences in theanalysis methods employed, and the dependent measuresused, as well as differences in the selection of brain regionsto investigate, in addition to differences in the post-injurytime of the study, and differences in whether subjects hadpositive or negative findings on conventional CT or MRI, itis surprising that there is as much convergence and consis-tency with respect to the detection of brain abnormalities inmTBI using DTI. Given the DTI findings reviewed above,and listed in Tables 3 and 4, as well as the frequently ofnegative CT and structural MRI scans, it is also quite clearthat DTI is by far the most sensitive in vivo method to detectsubtle brain abnormalities in mTBI.
While each of the 43 DTI studies investigating mTBI,and included in Tables 3 (n032) and 4 (n011), report someDTI abnormalities, their anatomical location does not al-ways converge. This lack of convergence is not, however,surprising, given the heterogeneity of brain injuries, as wellas the variability in these studies between time of injury andDTI scan. Some regions, however, are reported more oftenthan the others, which might further suggest their increasedvulnerability to axonal injury.
For example, in reviewing specific brain regions, some,as noted above, including the corpus callosum and parts ofthe corpus, are abnormal in mTBI (e.g., Arfanakis et al.2002; Bazarian et al. 2007; Grossman et al. 2011; Henry etal. 2011; Holli et al. 2010b; Huisman et al. 2004; Inglese etal. 2005; Kumar et al. 2009; Lipton et al. 2008; Little et al.
2010; Ljungqvist et al. 2011; Lo et al. 2009; Messe et al.2011; Maruta et al. 2010; Matthews et al. 2011; Mayer et al.2010; McAllister et al. 2012; Miles et al. 2008; Niogi et al.2008a; b; Rutgers et al. 2008a; b; Singh et al. 2010; Smits etal. 2011; Warner et al. 2010(b); Yurgelun-Todd et al. 2011).
Rutgers and coworkers (2008a) have also reported re-duced FA predominantly in 9 major regions in mTBIpatients, and one of these regions included the corpus cal-losum. In a separate study by this group of investigators(Rutgers et al. 2008b), mTBI and moderate TBI subjectswere included. Patients with mTBI or less than 3 monthspost-injury showed reduced FA and increased ADC in thegenu of the corpus callosum, whereas patients with greaterthan 3 months post-injury showed no such differences. Thisgroup of investigators also showed that more severe traumawas associated with FA reduction in the genu and spleniumof the corpus callosum, along with increased ADC andfewer numbers of fibers. These findings suggest that theremay be a reversal of damage to the corpus callosum inpatients with mTBI who recover after 3 months, and thatmore severe damage to the corpus callosum may be associ-ated with a worse outcome. Again, a longitudinal study thatinvestigates the course of injury over time would provideimportant information regarding the staging, progression,and possible recovery and reversal of brain injuries overtime.
Huisman et al. (2004) included TBI patients with Glas-gow Coma Scale scores between 4 and 15, and hencemoderate and severe were not separated from mild in eval-uating the corpus callosum. Matthews et al. (2011) exam-ined mTBI patients with major depressive disorder andthose without a major depressive disorder. These investiga-tors found that those with a major depressive disordershowed reduced FA in the corpus callosum (also coronaradiata and superior longitudinal fasciculus) compared withthose without a major depression. Zhang et al. (2010), onthe other hand, reported no differences in the corpus cal-losum using either whole brain analysis or region of interestanalyses between mTBI athletes and controls, but mTBIpatients did show greater variability of FA in the genu andin the body of the corpus callosum than did controls. Langeet al. (2011) also reported no differences in FA or MDbetween mTBI and controls, although they did report anon-significant trend for an increase in MD in the spleniumof the corpus callosum in mTBI compared with controls.MacDonald et al. (2011) also reported no differences ingenu or splenium of corpus callosum between mTBI andcontrols. The controls in this study, however, were 21 con-trols with blast exposure but without head trauma. It may be,however, that exposure to blasts results in subconcussiveinjury to the brain as is reported in sports-related injuries(e.g., Cubon and Putukian 2011; McAllister et al. 2012).Accordingly, fewer findings between groups might be
180 Brain Imaging and Behavior (2012) 6:137–192
accounted for by the fact that the control group is moresimilar to the mTBI group. Levin et al. (2010) also didnot find differences for the corpus callosum between TBIpatient and controls on FA or ADC measures, but heremTBI cases were not separated from moderate TBI andthe controls were 15 veterans, 7 with extracranial injury.
Again, the issue of using a military population as acontrol group needs more careful consideration as many ofthese individuals may have sustained blast injuries wherethey did not experience loss of consciousness but whichmay, nevertheless, have resulted in subtle alterations to thebrain (and thus use of such controls might result in a de-crease in the sensitivity of the imaging methods).
Davenport and coworkers (2011) selected veterans withand without blast exposure in their study where they investi-gated 25 veterans returning from Operation Enduring Free-dom and Operation Iraqi Freedom who had been exposed toblasts and subsequently developed symptoms indicative ofmTBI, and they included a comparison group comprised of33 veterans who had not experienced either blast exposure orhead injury during their tour of duty. These investigatorsreported diffuse and global patterns of reduced FA in whitematter in the blast exposure group compared to the group notexposed to blast. Furthermore, those who experienced morethan one blast related mTBI showed a larger number ofreduced FAvoxels in the brain. Interestingly, and surprisingly,58 % of the no blast group had experienced a previous mTBIas civilians, prior to entering the service. This group did showdifferences, however, compared with the blast exposure groupwith symptoms. Here the issue may not be exposure per se,but the differentiation between those with and without symp-toms being more important in differentiating between groups.Yurgelun-Todd et al. (2011) also investigated a sample of 15veterans with mild, moderate, and severe TBI and comparedthem with 10 civilians and 6 veterans. They reported de-creased FA in left cingulum and genu of the corpus callosum.The TBI group showed greater impulsivity. In addition, bothtotal and right cingulum FA reduction was correlated withincreased suicidal ideation and increased impulsivity.
Other brain regions reported as abnormal in mTBI in-clude the internal capsule (Arfanakis et al. 2002; Bazarian etal. 2007; Bazarian et al. 2011; Cubon and Putukian 2011;Grossman et al. 2011; Huisman et al. 2004; Inglese et al.2005; Lipton et al. 2008; Lo et al. 2009; Mayer et al. 2010;Miles et al. 2008), the external capsule (Arfanakis et al.2002; Bazarian et al. 2007; Bazarian et al. 2011), the cen-trum semiovale (Grossman et al. 2011; Holli et al. 2010b;Inglese et al. 2005; Miles et al 2008), inferior fronto-occipital fasciculus (Cubon and Putukian 2011; Messe etal. 2011; Singh et al. 2010; Smits et al. 2011), inferiorlongitudinal fasciculus (Cubon and Putukian 2011; Messeet al. 2011; Niogi et al. 2008a; b; Singh et al. 2010), superiorlongitudinal fasciculus (Cubon and Putukian 2011; Geary et
al. 2010; Mayer et al. 2010; Matthews et al. 2011; Niogi etal. 2008a; b), uncinate fasciculus (Geary et al. 2010; Mayeret al. 2010; Niogi et al. 2008a; b; Singh et al. 2010), coronaradiata (Little et al. 2010; Maruta et al. 2010; Matthews et al.2011; Mayer et al. 2010; Niogi et al. 2008a; b), corticospinaltract (Henry et al. 2011; Messe et al. 2011; Singh et al.2010), the cingulum bundle (MacDonald et al. 2011; Niogiet al. 2008a; b; Rutgers et al. 2008a), forceps minor andmajor (parts of corpus callosum; Messe et al. 2011), forcepsmajor (Little et al. 2010; Messe et al. 2011), cerebral lobarwhite matter (Lipton et al. 2009; Rutgers et al. 2008a),mesencephalon (Hartikainen et al. 2010; Holli et al. 2010b),the sagittal stratum (Cubon and Putukian 2011; Geary et al.2010; ), frontal lobe, parietal lobe, temporal lobe, occipitallobe (Salmond et al. 2006), dorsolateral prefrontal cortex(Lipton et al. 2009; Zhang et al. 2010), cerebellar peduncles(MacDonald et al. 2011; Maruta et al. 2010), hippocampus(Singh et al. 2010 ), fornix (Singh et al. 2010 ), thalamus(Grossman et al. 2011), thalamic radiation (Cubon andPutukian 2011; Messe et al. 2011), orbitofrontal white matter(MacDonald et al. 2011), subcortical white matter (Liptonet al. 2008), fronto-temporo-occipital association fiber bun-dles (Rutgers et al. 2008a), acoustic radiation (Cubon andPutukian 2011), and deep cortical brain regions (Brandstacket al. 2011).
In summary, there is a great deal of variability in DTIstudies of mTBI with respect to the time period of scanningat post-injury, as well as other factors including magnetstrength, brain regions examined, and methods of analyses,as noted above. Further, differences in both anatomicallocation of observed brain alterations, as well as in thenature of these alterations (i.e., increased versus decreasedFA), all contribute to the difficulty in interpreting this bodyof data. The findings are nonetheless striking in that they allsuggest that radiological evidence supports small and subtlebrain injuries in mTBI (see Table 3 and 4 for details). Thisevidence would not be possible if conventional MRI and CTscans alone were used to establish brain injury; it takes moreadvanced and sophisticated methods such as DTI that aresensitive to diffuse axonal injury to delineate these abnor-malities. Most of the studies reviewed, however, were cross-sectional studies, with scans acquired at different times post-injury. The cross-sectional nature of most of these studiesfurther highlights the need for longitudinal studies thatfollow a group of mTBI patients from early in the courseof injury, i.e., the first 24 to 72 h, with repeat scans at severalweeks, 1 month, 3 months, 6 months, and 1 year. Such datawould provide important information regarding the reversalof damage as well as provide important information aboutwhat radiological features early in the course of injurypredict good outcome versus poorer outcome with concom-itant post-concussive symptoms. There is also a clear needfor individualized imaging biomarkers (personalized
Brain Imaging and Behavior (2012) 6:137–192 181
medicine approach), where a single subject could be com-pared to a normative group for accurate diagnosis, with MRbiomarkers also providing information about prognosis inorder to implement treatment plans. These and other issuesare the topic of future directions of research, which follows.
Future directions of research
Summary Until quite recently no sufficiently robust radio-logical technologies existed, on either the acquisition orpost-processing side, to detect small and subtle brain alter-ations that are extant in mTBI (see also Irimia et al. 2011).Today, there are advanced image acquisition sequencesavailable to identify and to quantify small regions of extra-and intra-cortical bleeding (i.e., SWI), as well as diffuseaxonal injuries (i.e., DTI). These advanced imaging toolsare particularly important for investigating brain pathologyin mTBI, where conventional MRI and CT imaging havefailed to provide radiological evidence of brain injuries.Hence we are now able to detect and to localize brainalterations in mTBI, which is a critical first step, as it meansthat the diagnosis of mTBI can be based on radiologicalevidence, rather than symptoms alone. We are now also ableto evaluate the extent of brain injuries. Moreover, we havethe capability to observe changes post-injury, and longitu-dinally, to determine prognosis based on early and interme-diate stages of brain injury. The latter may lead to the earlyidentification of those individuals who are likely to recoveryversus those who are more likely to experience prolongedpostconcussive symptoms.
Thus a focus on longitudinal studies is needed to under-stand the dynamic nature of brain injury changes over time,as there is very little information about the different patternsof recovery versus non-recovery, and how these changes arereflected by a given imaging modality. For example, lesionsmay be visible using one imaging modality but not another,depending upon the pattern and stage of recovery. Conse-quently, the combined use of multimodal imaging, usingsemi-automated measures for analyses, and including a lon-gitudinal focus would greatly further our understanding ofmTBI. These and other future directions for research areelaborated upon below.
Multi-modal imaging and standardized protocols What isneeded is the establishment of acquisition protocols thatinclude multimodal imaging to characterize brain alterationsin mTBI, as one imaging modality may not accuratelycapture brain alterations in mTBI. Both Duhaime and cow-orkers (2010) and Haacke and coworkers (2010) also dis-cuss the need for more standardized image acquisitionprotocols for both research and for clinical purposes. It isimportant also to select optimal imaging acquisition
protocols that characterize the kind of injury that is presentin mTBI. This approach has been used successfully inresearch in Alzheimer’s disease, multiple sclerosis, and inpsychiatric disorders such as schizophrenia (e.g., Leung etal. 2010; Zou et al. 2005). Following these principles is theClinical Consortium on PTSD and TBI (INTRuST; seedetails: http://intrust.spl.harvardu.edu/pages/imaging andsee http://intrust.sdsc.edu/), a new initiative sponsored bythe Department of Defense, which includes the acquisitionof multimodal imaging, from six multi-national centers,using a high resolution, multi-shell DTI protocol as well aSWI and resting-state fMRI acquisition sequences, in addi-tion to anatomical MRI sequences.
More specific measures of microstructural abnormalities Withregard to diffusion imaging, future advances will likelyinclude more specific parameters than FA and MD. Forexample, free-water is a measure that can be derived fromconventional DTI to increase the specificity of observedmicrostructural abnormalities. The free-water model (Pasternaket al. 2009) assumes that there are two distinct compart-ments, one that is comprised of water molecules that arefreely diffusing in extracellular space, and another that iscomprised of the remaining molecules, which are restrictedby the cellular tissue. As a result, it is possible to estimateexplicitly the volume of extracellular water, a measure thatis sensitive to vasogenic edema, and therefore consequentlylikely specific to neuroinflammation as well. In addition,DTI indices such as FA and MD that are corrected for free-water can be used to estimate the properties of the remainingcompartment, i.e., cellular tissue. These corrected values aretissue specific (Pasternak et al. 2009, 2011c) and the corre-lation between corrected FA and corrected MD is reduced(Metzler-Baddeley et al. 2012), which makes it possible touse these two parameters, separately, to evaluate tissuedegeneration versus extracellular swelling (vasogenic ede-ma and inflammation) and cell swelling (cytotoxic edema).In the very near future, this free-water method may beutilized to investigate and to track brain injury over timeand to establish possible biomarkers that predict outcomebased on knowing the type (vasogenic versus cytotoxic),location, and extent of the edema.
Multi-shell diffusion weighted imaging (DWI) and kurtosisMulti-shell diffusion imaging is another new approach thatmay be useful in future imaging studies of mTBI. Here,DWI is acquired at multiple b-values in the same sessionand it provides additional microstructural information aboutthe organization of white and gray matter. In these scansequences, b-values, which are higher than the standardDTI acquisition (e.g., 3000 mm/s2), are used to obtaindiffusion properties that vary with different gradientstrengths. Diffusion Kurtosis Imaging (DKI), a technique
182 Brain Imaging and Behavior (2012) 6:137–192
that uses multi-shell diffusion imaging, measures the non-Gaussian behavior of water diffusion, and has been shownto be sensitive to characterizing subtle changes in neuronaltissue (Jensen et al. 2005).
The kurtosis measure is likely more sensitive to biologicalprocesses such as “reactive astrogliosis” compared with themore general measures of FA and MD (Zhou et al. 2012).Specifically, kurtosis is a measure of the deviation from thediffusion tensor model (Gaussian diffusion), and thus comple-ments the other measures that are derived from the tensormodel. For example, the directionally specific kurtosis mea-sure has been shown to be more sensitive to developmentalchanges in white and gray matter than FA or MD in rat brains(Cheung et al. 2009). Thus, in this way, kurtosis is moresensitive than traditional diffusion measures to changes inrestricted diffusion and thereby has the potential to detectchanges related to cells (i.e., astrocytes, neurons, andoligodendrocytes), as opposed to less specific measures ofchanges in the integrity of tissue (e.g., FA and MD).
Lack of homogeneity in brain trauma and the need for apersonalized medicine approach While we have onlytouched the surface with respect to what more sensitiveimaging technology can inform us about mTBI, another areaof interest for future research is to go beyond group analy-ses, i.e., comparing a group of patients with mTBI with agroup of normal controls. Group analyses, in fact, are some-what misleading because they are predicated on findinglarger differences between groups than within groups. Un-fortunately, brain trauma is a very heterogeneous disorder,and group analyses, particularly those that use whole brainanalyses (i.e., comparing average mTBI brain to averageMRI brain), are not well suited for accurate analysis becausethey obscure the individual differences that characterizebrain injuries. The heterogeneity of brain trauma is a topicthat is also reviewed in this issue by Ms. Sara Rosenbaumand Dr. Michael Lipton.
To circumvent this heterogeneity, a personalized ap-proach needs to be developed that can be incorporated intocomparisons between mTBI and normal controls. Such anapproach must go beyond a visual inspection of DTI mapssuch as FA and ADC, to a quantitative, objective approach.The latter may then be used as both radiological evidence ofbrain injury and to provide an individual profile of braininjury, as well as to form the basis for comparing normalcontrols and patients with mTBI.
One possible approach to this problem is to build norma-tive atlases based on normal controls, along with robuststatistics (e.g., z-score and receiver operator-ROC- curves)to detect “out-of-the-normal” imaging features (e.g., FA,MD, free-water, and kurtosis) at various locations in thebrain. This information may then be used to detect brainregions that are abnormal in any of these measures, leading
to a diagnosis of diffuse axonal injury (FA, MD), or edema(free water), or myelin damage (kurtosis), which would bespecific for each individual case. In this way, one could notonly establish a profile of injury for each individual patient,but one could also perform group analyses based on theseverity and load of injuries (i.e., extent and number of brainregions involved), without relying on the assumption of acommon pattern of injury location among all patients withTBI. The latter assumption is the current approach to groupanalyses including brain-wise, voxel-based analyses thatcreate averages of brain regions, i.e., comparing a group ofhealthy and injured brains simultaneously, and demonstrat-ing where in the brain for the entire group of mTBI (not eachindividual), differs from a group of healthy controls.
Figure 8 shows an example of one mTBI case comparedto a normative atlas of FA and of MD, where z-score mapshighlight brain regions that are more than three standarddeviations from the normative brain atlas for FA (top offigure) and MD (bottom of figure). This is the kind ofinformation that can be evaluated for individual cases,where normative atlases can be built to examine any numberof measures (i.e., FA, MD, free-water, kurtosis, etc.).
A further example is shown in Figure 9, which depictsmore than three standard deviations from the normativebrain atlas for free-water, where subject specific maps werecreated to show abnormal free-water (edema) in two indi-viduals. The individual on the top shows more localizedabnormalities, while the individual on the bottom showsmore small, scattered lesions that have an edematous orneuroinflammatory component (Pasternak et al. 2011b).These findings may vary across subjects, from localizedpathologies to having a more diffuse and scattered patternof pathology, as seen here, and, importantly, provide infor-mation about individual pathology that is not detected usingconventional MR or CT. These variations pose extremechallenges to group comparisons, though group compari-sons may still be made. They also highlight the heterogene-ity that is characteristic of TBI, including mTBI.
Tractography Another promising method for probing alter-ations in white matter in mTBI is tractography. The wordtractography refers to any method for estimating the trajec-tories of the fiber tracts (bundles) in the white matter. Manymethods have been proposed for tractography, and theresults will vary depending on the chosen method. Forexample, deterministic tractography involves directly fol-lowing the main diffusion direction, whereas probabilisticmethods estimate the likelihood of two regions beingconnected (Bjornemo et al. 2002; Behrens et al. 2003).The most common deterministic approach is streamlinetractography (see Figure 10) (Basser et al. 2000; Conturoet al. 1999; Mori et al. 1999; Westin et al. 1999), which isclosely related to an earlier method for visualization of
Brain Imaging and Behavior (2012) 6:137–192 183
tensor fields known as hyperstreamlines (Delmarcelle andHesselink 1992). For a clinical and technical overview oftractography in neurological disorders, the reader is referredto Ciccarelli et al. (2008). For reviews of tractographytechniques including explanations of common tractographyartifacts and a comparison of methods, the reader is referredto Jones (2008) and Lazar (2010).
DTI based streamline tractography has been used by severalinvestigative groups to examine schizophrenia and other neu-rological and psychiatric disorders (see review in Kubicki et al.2007; Shenton et al. 2010; Whitford et al. 2011). The mainlimitation of this method is that it does not allow one to tracefibers through complex fiber orientations such as branchingand crossing fibers. This is because DTI based streamline
tractography assumes that there exists only one fiber bundleoriented coherently in a single direction at each voxel. Thusvoxels where different fiber bundles cross cannot be charac-terized using this model. Consequently, several advancedmod-els have been proposed in the literature, such as a high-ordertensor model (Ozarian 2003; Barmpoutis et al. 2009), a spher-ical harmonics based nonparametric model (Anderson 2005)and a multi-tensor model (Tuch et al. 2002; Pasternak et al.2008; Malcolm et al. 2010), among others.
Figure 10 shows a method of tracing the corpus callosumfiber bundle using single tensorDTI based tractography (on thetop) and another method using two-tensor DTI based tractog-raphy (on the bottom) that illustrates the improved fiber track-ing using the two-tensor method. This method makes it
Fig. 8 Z-score maps for apatient with chronic mTBIsubject. Z-score maps werecreated from a comparison toa normative atlas. The regionsin red and yellow showstatistically significantabnormal regions for eitherFA (top) or MD (bottom)
Fig. 9 Z-score maps for twopatients with chronic mTBI.Z-score maps were created froma comparison to a normativeatlas. The images at the topshow a patient with morelocalized regions of increasedfree-water while the images atthe bottom show a patientwith more diffuse regions ofincreased free-water (blue colorindicates statistically significantareas of free-water comparedwith the normative atlas)
184 Brain Imaging and Behavior (2012) 6:137–192
possible now to trace multiple fiber tracts that traverse the brainand connect several different brain regions (e.g., Malcolm et al.2010; Pasternak et al. 2008; Tournier et al. 2008).
Figure 11 shows all of the fibers in the brain that travelthrough the corpus callosum, in a single case, using two-tensor tractography (Malcolm et al. 2010). Of note here, themethodology for analysis can remain the same as before, i.e.,trace a certain fiber bundle in healthy controls and compute the“typical” diffusion properties such as FA, MD, etc., then tracethe same bundle in TBI subjects, and finally compare them todetect differences. What is different in terms of moving from asingle tensor model to two and multi-tensor models is that itimproves the ability to accurately identify crossing fiber tracts.
Mean-squared-displacement and return-to-origin probabilitiesOther more sensitive markers of diffusion, such as mean-
squared-displacement and return-to-origin probabilities, canbe obtained from more advanced diffusion scans such asmulti-shell (multiple b-values) diffusion imaging (e.g.,Mitra 1992; Yu-Chien et al. 2007). Although the scan timefor these acquisitions is longer, they provide more subtleinformation beyond what can be obtained from DTI as it iscurrently used. For example, the return-to-origin probabilitymeasures the probability that a water molecule will return toits starting point in a given amount of time. This measurewill be higher for white matter due to restriction on themotion of water as a result of the coherent layout of thewhite matter fibers. This probability will be lower in graymatter and lowest in CSF, where water is essentially free tomove. Thus this measure captures subtle changes in theorganization of white matter and has been shown to be moresensitive to de-myelination of white matter tracts comparedto DTI measures such as FA (Assaf et al. 2005).
Another diffusion measure that is more sensitive to therestriction of water molecules due to cellular boundaries ismean-square-displacement (MSD)(e.g., Assaf et al. 2002).This measure is the mean distance travelled by a watermolecule in a given diffusion time. Thus, MSD is highly
1-Tensor
2-Tensor Kalman Filter
Fig. 10 Part of the corpus callosum fibers, where seeding was done inthe mid-sagittal plane of the corpus callosum. Top figure shows thetracing using the standard single-tensor model and bottom figureshows tracts generated with the two-tensor model
Fig. 11 Panel A shows a coronal view of white matter fiber tractsusing a two-tensor model that go through the corpus callosum Panel Bis a sagittal view of the corpus callosum shown in Panel A
Brain Imaging and Behavior (2012) 6:137–192 185
dependent on the cellular structure at any given location andhence could be related to the concentration of certain bio-chemicals. This could be validated by correlating this mea-sure with the concentration of certain biochemicals such asphospholipids as obtained using an MRS scan. Hence, suchmeasures can be very useful for detecting mTBI since theunderlying tissue changes are often very subtle and may bemissed by current DTI and DKI imaging protocols. In thenear future, this approach could also be combined withother brain biomarkers that are based, for example, onblood-serum, the latter a powerful complementary toolthat provides information about genes and proteins af-fected by brain injury (see section below; see alsoMondello et al. 2011).
Identification and delineation of brain injury, and thedevelopment of biomarkers for possible treatment andtreatment efficacy trials Advanced multi-modal neuroimag-ing techniques provide radiological evidence of mTBI thatgo beyond self-report and other more conventional meas-ures, and may elucidate further the mechanisms and neuro-physiology underlying mTBI. This is an important first step.More studies, however, are needed using multi-modal im-aging techniques. Further, based on their enhanced sensitiv-ity, many of these more advanced techniques should be usedto monitor treatment efficacy, and serve as endpoints fornew trials of medication aimed at neuroplasticity or neuro-inflammation in TBI.
One example of multi-modal neuroimaging is to com-bine MRS and DTI in the same subjects. MRS and DTIare highly complimentary given the biochemical andstructural focus of each modality. However, few studieshave utilized the two together. In severe head injury, thecombination of these two modalities has provided greaterdiagnostic accuracy for predicting outcome one year fol-lowing injury (Tollard et al. 2009), than either MRS orDTI alone. It is likely that this same combination wouldalso provide greater sensitivity to the more subtle changes inmTBI.
The co-localization of DTI and MRS may also providegreater insight into the underlying physiological changesthat occur in mTBI. An early MRS study by Cecil andcoworkers (Cecil et al. 1998) did not have DTI available atthe time, but their findings of decreased N-acetyl aspartate(NAA; a putative neuronal and axonal marker; see review inthis issue by Lin et al.) in the corpus callosum corroboratesDTI findings of reduced FA, as previously described in thecurrent review. As NAA is transported down axons and theloss of FA is attributed to axonal injury, a strong correlationbetween these two measures further strengthens the argu-ment for axonal injury. NAA MRS measures can also helpto validate DTI studies in regions of fiber crossings andconfirm that decreased FA can be attributed to loss of fiber
integrity as opposed to technical issues. Studies of schizo-phrenia (e.g., Tang et al. 2007) and motor neuron disease(Nelles et al. 2008) have also utilized MRS and DTI in thiscomplimentary fashion.
Additionally, by combining different biomarkers fromblood with MR biomarkers, we may be able to delineatefurther specific types of brain injury involved in mTBI, aswell as how they change over time, and what the evolutionis of secondary damage or progression of types of injuryover time. More specifically, serum biomarker profiles, asan outcome measure of brain damage, are being used todetect brain damage in ischemic stroke and TBI in animalstudies (Liu et al 2010). This approach may also be useful inhumans where TBI-specific biomarkers could be developed,based on proteomics, to provide distinctive profiles of spe-cific genes and proteins that are altered in mTBI (e.g.,Mondello et al. 2011). The free-water model, describedabove, could also be combined with measures of proteinsinvolved in neuroinflammation in the brain to develop com-bined TBI-specific biomarkers that may detect brain injuriesand predict course, outcome, and treatment response betterthan using either a proteomic or MR biomarker alone.
Thus combining multimodal imaging modalities, andcombining genomic and proteomic biomarkers with MRbiomarkers, may lead to the discovery of more specificbiomarkers of biochemical and physiological processes,which, in turn, may provide important new informationabout primary and secondary consequences of injury, andassist in determining what combination of biomarkers areindicative of axonal injury, inflammation, demyelination,apoptosis, neuroregeneration, etc., and what combinationbest predict outcome and treatment. It is also likely thatthe development of multiple biomarkers will lead to a newstratification of patients based on several biomarkers that,when combined, are more specific to important measures ofbrain injury than is any one biomarker alone. This newapproach to developing multi-modal MR biomarkers andcombining these with serum based proteomic biomarkersto investigate brain injuries in mTBI may go a long way toproviding more accurate diagnosis of mTBI, as well as toproviding important indicators of treatment response andoutcome.
Acknowledgements This work was supported in part by theINTRuST Posttraumatic Stress Disorder and Traumatic Brain InjuryClinical Consortium funded by the Department of Defense PsychologicalHealth/Traumatic Brain Injury Research Program (X81XWH-07-CC-CS-DoD; MES, JSS, SB, OP, MK, YR, M-AV, C-FW, RZ), by an NIHNINDS funded R01 (R01 NS 078337; RS, MES, JSS), by a Center forIntegration of Medicine (CIMIT) Soldier in Medicine Award (SB, MES),by an NIH NIMH funding R01 (R01 MH082918; SB), by funding fromthe National Research Service Award (T32AT000051; MPP) from theNational Center for Complementary and Alternative Medicine(NCCAM) at the National Institute of Health, by the Harvard MedicalSchool Fellowship as part of the Eleanor and Miles Shore Fellowship
186 Brain Imaging and Behavior (2012) 6:137–192
Program (HH), by the Deutsche Akademischer Austauschdienst(DAAD; IK), and by funding from NCRR, including the NationalAlliance for Medical Image Computing (NAMIC-U54 EBOO5149;RK, MK, MES), and the Neuroimaging Analysis Center (NAC;P41RR13218; RK and CF).
References
Adams, J. H., Doyle, D., et al. (1989). Diffuse axonal injury inhead injury: definition, diagnosis, and grading. Histopathology,15, 49–59.
Alexander, M. P. (1995). Mild TBI: pathophysiology, natural history,and clinical management. Neurology, 45(7), 253–260.
Anderson, A. (2005). Measurement of fiber orientation distributionsusing high angular resolution diffusion imaging. Magnetic Reso-nance in Medicine, 54(5), 1194–1206.
Anderson, C. V., Bigler, E. D., et al. (1995). Frontal lobe lesions,diffuse damage, and neuropsychological functioning in traumaticbrain-injured patients. Journal of Clinical and ExperimentalNeuropsychology, 17, 900–908.
Anderson, C. V., Wood, D. M., et al. (1996). Lesion volume, injuryseverity, and thalamic integrity following head injury. Journal ofNeurotrauma, 13(2), 59–65.
Arciniegas, D. B., Anderson, C. A., et al. (2005). Mild traumatic braininjury: a neuropsychiatric approach to diagnosis, evaluation, andtreatment. Neuropsychiatric Disease and Treatment, 1, 311–327.
Arfanakis, K., Haughton, V. M., et al. (2002). Diffusion tensor MRimaging in diffuse axonal injury. AJNR. American Journal ofNeuroradiology, 23(5), 794–802.
Ashwal, S., Holshouser, B. A., et al. (2006). Use of advanced neuro-imaging techniques in the evaluation of pediatric traumatic braininjury. Developmental Neuroscience, 28, 309–326.
Assaf, Y., & Pasternak, O. (2008). Diffusion tensor imaging (DTI)-based white matter mapping in brain research: a review. Journalof Molecular Neuroscience, 34(1), 51–61.
Assaf, Y., Ben-Bashat, D., et al. (2002). High b-value q-space analyzeddiffusion-weighted MRI: application to multiple sclerosis.Magnetic Resonance in Medicine, 47(1), 115–126.
Assaf, Y., Chapman, J., Ben-Bashat, D., et al. (2005). White matterchanges in multiple sclerosis: correlation of q-space diffusion MRIand 1-H MRS. Magnetic Resonance Imaging, 23(6), 703–710.
Babikian, T., Freier, M. C., et al. (2005). Susceptibility weightedimaging: neuropsychologic outcome and pediatric head injury.Pediatric Neurology, 33(3), 184–194.
Babikian, T., Freier, M. C., et al. (2006). MR spectroscopy: predictinglong-term neuropsychological outcome following pediatric TBI.Journal of Magnetic Resonance Imaging, 24, 801–811.
Barkhoudarian, G., Houda, D. A., et al. (2011). The molecular patho-physiology of concussive brain injury. Clinics in Sports Medicine,30, 33–48.
Barmpoutis, A., Hwang, M. S., Howland, D., et al. (2009). Regularizedpositive-definite fourth order tensor field estimation from DW-MRI. NeuroImage, 45(1), S153–S162.
Basser, P. J., Mattiello, J., et al. (1994). MR diffusion tensor spectros-copy and imaging. Biophysical Journal, 66(1), 259–267.
Basser, P. J., Pajevic, S., Pierpaoli, C., et al. (2000). In vivo fibertractography using DT–MRI data. Magnetic Resonance inMedicine, 44, 625–632.
Bazarian, J. J., Wong, T., et al. (1999). Epidemiology and predictors ofpost-concussive syndrome after minor head injury in an emergen-cy population. Brain Injury, 13(3), 173–189.
Bazarian, J. J., Blyth, B., et al. (2006). Bench to bedside: evidence forbrain injury after concussion–looking beyond the computed to-mography scan. Academic Emergency Medicine, 13(2), 199–214.
Bazarian, J. J., Zhong, J., et al. (2007). DTI detects clinically importantaxonal damage after mild TBI: a pilot study. Journal of Neuro-trauma, 24(9), 1447–1459.
Bazarian, J. J.,& Zhu, T., et al. (2011). Subject-specific changes in brainwhite matter on diffusion tensor imaging after sports-related concus-sion. Magnetic Resonance Imaging. Nov 11. [Epub ahead of print]
Behrens, T., Woolrich, M., Jenkinson, M., et al. (2003). Characteriza-tion and propagation of uncertainty in diffusion-weighted MRimaging. Magnetic Resonance in Medicine, 50, 1077–1088.
Belanger, H. G., Vanderploeg, R. D., et al. (2007). Recent neuroimag-ing techniques in mild traumatic brain injury. The Journal ofNeuropsychiatry and Clinical Neurosciences, 19(1), 5–20.
Belanger, H. D., Kretzmer, T., et al. (2009). Symptom complaintsfollowing combat-related traumatic brain injury: Relationshipto traumatic brain injury severity and posttraumatic stressdisorder. Journal of the International NeuropsychologicalSociety, 16, 194–199.
Bell, R. S., & Loop, J. W. (1971). The utility and futility of radio-graphic skull examination for trauma. The New England Journalof Medicine, 284(5), 236–239.
Benson, R. R., Meda, S. A., et al. (2007). Global white matter analysisof diffusion tensor images is predictive of injury severity in TBI.Journal of Neurotrauma, 24(3), 446–459.
Bergeson, A. G., Lundin, R., et al. (2004). Clinical rating of corticalatrophy and cognitive correlates following traumatic brain injury.Clinical Neuropsychology, 18(4), 509–520.
Bigler, E. D. (2004). Neuropsychological results and neuropathologicalfindings at autopsy in a case of mild traumatic brain injury. Journalof the International Neuropsychological Society, 10(5), 794–800.
Bigler, E. D. (2008). Neuropsychology and clinical neuroscience ofpersistent post-concussive syndrome. Journal of the InternationalNeuropsychological Society, 14(1), 1–22.
Bigler, E. D., Blatter, D. D., et al. (1997). Hippocampal volume innormal aging and traumatic brain injury. AJNR. American Journalof Neuroradiology, 18(1), 11–23.
Bjornemo, M., Brun, A., Kikinis, R., & Westin, C.-F. (2002). Regu-larized stochasticwhite matter tractography using diffusion tensorMRI. Medical Image Computing and Computer Assisted Inter-vention (MICCAI), 435–442.
Blumbergs, P. C., Scott, G., et al. (1994). Staining of amyloid precursorprotein to study axonal damage in mild head injury. Lancet, 344(8929), 1055–1056.
Bouix, S., Pelavin, P., et al. (2011). Diagnosis of diffuse axonal injurywith diffusion tensor imaging. The 3rd Federal Interagency Con-ference on TBI, Washington DC.
Brandstack, N., Kurki, T., et al. (2011). Diffusivity of normal-appearing tissue in acute traumatic brain injury. Clinical Neuro-radiology, 21(2), 75–82.
Brooks, W. M., Friedman, S. D., et al. (2001). Magnetic resonancespectroscopy in TBI. The Journal of Head Trauma Rehabilitation,16(2), 149–164.
Bruns, J. J., & Jagoda, A. (2009). Mild traumatic brain injury. TheMount Sinai Journal of Medicine, 76, 129–137.
Budde, M. D., Kim, J. H., et al. (2007). Toward accurate diagnosis ofwhite matter pathology using diffusion tensor imaging. MagneticResonance in Medicine, 57(4), 688–695.
Budde, M. D., Janes, L., et al. (2011). The contribution of gliosis todiffusion tensor anisotropy and tractography following traumaticbrain injury: validation in the rat using Fourier analysis of stainedtissue sections. Brain, 134(8), 2248–2260.
Carroll, L. J., Cassidy, J. D., et al. (2004a). Prognosis for mild trau-matic brain injury: results of the WHO collaborating centre taskforce on mild traumatic brain injury. Journal of RehabilitationMedicine, 43(Suppl.), 84–105.
Carroll, L. J., Cassidy, J. D., et al. (2004b). Systematic search andreview procedures: results of the WHO collaborating centre task
Brain Imaging and Behavior (2012) 6:137–192 187
force on mild traumatic brain injury. Journal of RehabilitationMedicine, 43(Suppl.), 11–14.
CDC (2010). Injury, prevention, & control: traumatic brain injury.Center for Disease Control and Prevention. http://www.cdc.gov/traumaticbraininjury/statistics.html
Cecil, K. M., Hills, E. C., Sandel, M. E., et al. (1998). Proton magneticresonance spectroscopy for detection of axonal injury in thesplenium of the corpus callosum of brain-injured patients. Journalof Neurosurgery, 88(5), 795–801.
Cheung, M. M., Hui, E. S., et al. (2009). Does diffusion kurtosisimaging lead to better neural tissue characterization? A rodentbrain maturation study. NeuroImage, 45(2), 386–392.doi:10.1016/j.neuroimage2008.12.018.
Ciccarelli, O., Catani, M., Johansen-Berg, H., et al. (2008). Diffusion-based tractography in neurological disorders: concepts, applica-tions, and future developments. Lancet Neurology, 7, 715–727.
Cohen, B. A., Inglese, M., et al. (2007). Proton MR spectroscopy andMRI-volumetry in mild traumatic brain injury. AJNR. AmericanJournal of Neuroradiology, 28(5), 907–913.
Coles, J. P. (2007). Imaging after brain injury. British Journal ofAnaesthesia, 99(1), 49–60.
Conturo, T. E., Lori, N. F., Cull, T. S., et al. (1999). Tracking neuronalfiber pathways in the living human brain. Neurobiology, 96,10422–10427.
Cubon, V. A., & Putukian, M. (2011). A diffusion tensor imaging studyon the white matter skeleton in individuals with sports-relatedconcussion. Journal of Neurotrauma, 28(2), 189–201.
Davenport, N. D., Lim, K. O., et al. (2011). Diffuse and sptiallyvariable white matter disruptions are associated with blast-related mild traumatic brain injury. Neuroimage. Oct 20. [Epubahead of print]
Delmarcelle, T., &Hesselink, L. (1992). Visualization of second order tensorfields and matrix data. Proceedings IEEE Visualization, 316–323.
Ding, K., Marquez de Plata, C., et al. (2008). Cerebral atrophy aftertraumatic white matter injury: correlation with acute neuroimag-ing and outcome. Journal of Neurotrauma, 25, 1433–1440.
Duhaime, A. C., Gean, A. D., et al. (2010). Common data elements inradiologic imaging of traumatic brain injury. Archives of PhysicalMedicine and Rehabilitation, 91(11), 1661–1666.
Eisenberg, H. M., & Levin, H. S. (1989). Computed tomography andmagnetic resonance imaging in mild to moderate head injury. InH. S. Levin, H. M. Eisenberg, & A. L. Benton (Eds.), Mild headinjury (pp. 133–141). NY: Oxford University Press.
Ennis, D. B., & Kindlmann, G. (2006). Orthogonal tensor invariantsand the analysis of diffusion tensor magnetic resonance images.Magnetic Resonance in Medicine, 55(1), 136–146.
Fan, J., McCandliss, B. D., et al. (2005). The activation of attentionalnetworks. NeuroImage, 26(2), 471–479.
Faul, M. D., Xu, L., et al. (2010). TBI in the United States: Emergencydepartment visits, hospitalizations, and deaths 2002-2006.Atlanta (GA): Centers for Disease Control and Prevention,National Center for Injury Prevention and Control.
Fischl, B., van der Kouwe, A. C., et al. (2004). Automatically parcel-lating the human cerebral cortex. Cerebral Cortex, 14, 11–22.
Fitzgerald, D. B., & Crosson, B. A. (2011). Diffusion weightedimaging and neuropsychological correlates in adults with mildtraumatic brain injury. International Journal of Psychophysiology,82(1), 79–85.
Fujiwara, E., Schwartz, M. L., et al. (2008). Ventral frontal cortexfunctions and quantified MRI in traumatic brain injury. Neuro-psychologia, 46, 461–474.
Gale, S. D., Johnson, S. C., et al. (1995). Nonspecific white matterdegeneration following traumatic brain injury. Journal of theInternational Neuropsychological Society, 1(1), 17–28.
Gale, S. D., Baxter, L., et al. (2005). Traumatic brain injury and greymatter concentration: a preliminary voxel based morphometry
study. Journal of Neurology, Neurosurgery & Psychiatry, 76(7),984–988.
Garnett, M. R., Blamire, A. M., et al. (2000). Early proton magneticresonance spectroscopy in normal-appearing brain correlates withoutcome in patients following traumatic brain injury. Brain, 120(10), 2046–2054.
Geary, E. K., Kraus, M. F., et al. (2010). Verbal learning differences inchronic mild traumatic brain injury. Journal of the InternationalNeuropsychological Society, 16(3), 506–516.
Gentry, L. R. (1994). Imaging of closed head injury. Radiology, 191,1–17.
Gentry, L. R., Godersky, J. C., et al. (1988). MR imaging of headtrauma: review of the distribution and radiopathologic features oftraumatic lesions. AJR American Journal of Roentgenology, 150(3), 663–672.
Goldstein, M. (1990). Traumatic brain injury: a silent epidemic(Editorial). Annals of Neurology, 27, 327.
Green, R., Koshimori, Y., et al. (2010). Research digest. Understandingthe organic basis of persistent complaints in mTBI: findings fromfunctional and structural neuroimaging. NeuropsychologicalRehabilitation, 20(3), 471–478.
Greiffenstein, M. (2008). Clinical myths of forensic neuropsychology.Clin Neuropsych, 1–11.
Grossman, E. J., & Ge, Y., et al. (2011). Thalamus and cognitiveimpairment in mild traumatic brain injury: A diffusional kurtosisimaging study. Journal of Neurotrauma. Sep 15. [Epub ahead ofprint]
Haacke, E. M., Xu, Y., et al. (2004). Susceptibility weighted imaging(SWI). Magnetic Resonance in Medicine, 52, 612–618.
Haacke, E. M., Mittal, S., et al. (2009). Susceptibility-weighted imag-ing: technical aspects and clinical applications, part 1. AJNRAmerican Journal of Neuroradiology, 30(1), 19–30.
Haacke, E. M., Duhaime, A. C., et al. (2010). Common data elementsin radiologic imaging of traumatic brain injury. Journal ofMagnetic Resonance Imaging, 32(3), 516–543.
Hackney, D. B. (1991). Skull radiography in the evaluation of acute headtrauma: a survey of current practice. Radiology, 181(3), 711–714.
Hartikainen, K. M., Waljas, M., et al. (2010). Persistent symptoms inmild to moderate traumatic brain injury associated with executivedysfunction. Journal of Clinical and Experimental Neuropsychology,32(7), 767–774.
Hayes, R. L., & Dixon, C. E. (1994). Neurochemical changes in mildhead injury. Seminars in Neurology, 14, 25–31.
Henry, L. C., Tremblay, J., et al. (2011). Acute and chronic changes indiffusivity measures after sports concussion. Journal of Neuro-trauma, 28(10), 2049–2059.
Himanen, L., Portin, R., et al. (2005). Cognitive functions in relation toMRI findings 30 years after traumatic brain injury. Brain Injury,19, 93–100.
Hoge, C. W., McGurk, D., et al. (2008). Mild TBI in U. S. soldiersreturning from Iraq. The New England Journal of Medicine, 358(5), 453–463.
Hoge, C. W., Goldberg, H. M., et al. (2009). Care of war veterans withmild traumatic brain injury–flawed perspectives. The NewEngland Journal of Medicine, 360(16), 1588–1591.
Holli, K. K., Harrison, L., et al. (2010). Texture analysis of MRimages of patients with mild traumatic brain injury. BMCMedical Imaging, 10, 8.
Holli, K. K., Wäljas, M., et al. (2010). Mild traumatic brain injury:tissue texture analysis correlated to neuropsychological and DTIfindings. Academic Radiology, 17(9), 1096–1102.
Holshouser, B. A., Tong, K. A., et al. (2005). Proton MR spectroscopicimaging depicts diffuse axonal injury in children with TBI. AJNR.American Journal of Neuroradiology, 26(5), 1276–1285.
Huang, M., Theilmann, R., et al. (2009). Integrated imaging approachwith MEG and DTI to detect mild traumatic brain injury in
188 Brain Imaging and Behavior (2012) 6:137–192
military and civilian patients. Journal of Neurotrauma, 26(8),1213–1226.
Hughes, D. G., Jackson, A., et al. (2004). Abnormalities on magneticresonance imaging seen acutely following brain injury: correla-tion with neuropsychological tests and delayed recovery. Neuro-radiology, 46, 550–558.
Huisman, T. A., Schwamm, L. H., et al. (2004). Diffusion tensorimaging as potential biomarker of white matter injury in diffuseaxonal injury. AJNR. American Journal of Neuroradiology, 25(3),370–376.
Hunter, J. V., & Wilde, E. A., et al. (2011). Emerging imaging tools foruse with traumatic brain injury research. Journal of Neurotrauma.Oct 17. [Epub ahead of print]
Inglese, M., Makani, S., et al. (2005). Diffuse axonal injury in mildTBI: a diffusion tensor imaging study. Journal of Neurosurgery,103(2), 298–303.
Irimia, A., Chambers, M. C., et al. (2011). Comparison of acute and chronictraumatic brain injury using semi-automatic multimodal segmenta-tion of MR volumes. Journal of Neurotrauma, 11, 2287–2306.
Iverson, G. L., Lovell, M. R., et al. (2000). Prevalence of abnormal CT-scans following mild head injury. Brain Injury, 14(12), 1057–1061.
Jang, S. H. (2011). Diffusion tensor imaging studies on corticospinaltract injury following traumatic brain injury: a review. NeuroRehabilitation, 29(4), 339–345.
Jenkins, A., Teasdale, G., et al. (1986). Brain lesions detected bymagnetic resonance imaging in mild to severe head trauma. Lan-cet, 2, 445–646.
Jensen, J. H., Helpern, J. A., et al. (2005). Diffusional kurtosis imag-ing: the quantification of non-gaussian water diffusion by meansof magnetic resonance imaging. Magnetic Resonance in Medi-cine, 53(6), 1432–1440.
Johnston, K. M., Ptito, A., et al. (2001). New frontiers in diagnosticimaging in concussive head injury. Clinical Journal of SportMedicine, 11(3), 166–175.
Jones, D. K. (2008). Studying connections in the living human brainwith diffusion MRI. Cortex, 44, 936–952.
Kou, Z., Tong, K. A., et al. (2010). The role of advanced MR imagingfindings as biomarkers of TBI. The Journal of Head TraumaRehabilitation, 25(4), 267–282.
Kraus, M. F., Susmaras, T., et al. (2007). White matter integrity andcognition in chronic traumatic brain injury: a diffusion tensorimaging study. Brain, 130(10), 2508–2519.
Kubicki, M., McCarley, R., et al. (2007). A review of diffusion tensorimaging studies in schizophrenia. Journal of Psychiatric Re-search, 41(1–2), 15–30.
Kumar, R., Gupta, R. K., et al. (2009). Comparative evaluation ofcorpus callosum DTI metrics in acute mild and moderate traumaticbrain injury: its correlation with neuropsychometric tests. BrainInjury, 23(7), 675–685.
Kurtzke, J. F., & Kurland, L. T. (1993). The epidemiology of neuro-logic disease. In R. J. Joynt (Ed.), Clinical neurology, rev (Ch.66). Philadelphia: JB Lippincott.
Lange, R. T., & Iversion, G. L., et al. (2011). Diffusion tensor imagingfindings are not strongly associated with postconcussional disor-der 2 months following mild traumatic brain injury. Journal ofHead Trauma Rehabilitation. Jun 2. [Epub ahead of print]
Langlois, J. A., Rutland-Brown, W., et al. (2006). The epidemiologyand impact of traumatic brain injury: a brief overview. The Jour-nal of Head Trauma Rehabilitation, 21(5), 375–378.
Lazar, M. (2010). Mapping brain anatomical connectivity using whitematter tractography. NMR in Biomedicine, 23, 821–835.
Le Bihan, D. (1991). Molecular diffusion nuclear magnetic resonanceimaging. Magnetic Resonance Quarterly, 7(1), 1–30.
Le, T. H., & Gean, A. D. (2009). Neuroimaging of traumaticbrain injury. The Mount Sinai Journal of Medicine, 76(2),145–162.
Leung, K. K., Clarkson, M. J., et al. (2010). Robust atrophy ratemeasurement in Alzheimer’s disease using multi-site serial MRI:tissue-specific intensity normalization and parameter selection.NeuroImage, 50(2), 516–523.
Levin, H. S., Matiss, S., et al. (1984). Neurobehavioral outcomefollowing head injury: a three center study. Journal of Neurosur-gery, 66, 234–243.
Levin, H. S., Amparo, E., et al. (1987). Magnetic resonance imagingand computerized tomography in relation to the neurobehavioralsequelae of mild and moderate head injuries. Journal of Neuro-surgery, 66(5), 706–713.
Levin, H. S., Wilde, E., et al. (2010). Diffusion tensor imaging of mildto moderate blast-related traumatic brain injury and its sequelae.Journal of Neurotrauma, 27(4), 683–694.
Levine, B., Kovacevic, N., et al. (2008). The Toronto traumatic braininjury study: injury severity and quantified MRI. Neurology, 70,771–778.
Lin, A., Ross, B. D., et al. (2005). Efficacy of proton magneticresonance spectroscopy in neurological diagnosis and neurother-apeutic decision making. NeuroRx, 2(2), 197–214.
Lin, A. P., & Ramadan, S., et al. (2010). Neurochemical changes inathletes with chronic traumatic encephalopathy. In Radiologicalsociety of North America: Chicago, IL.
Lipton, M. L., Gellella, C., et al. (2008). Multifocal white matterultrastructural abnormalities in mild traumatic brain injury withcognitive disability: a voxel-wise analysis of diffusion tensorimaging. Journal of Neurotrauma, 25(11), 1335–1342.
Lipton, M. L., Gulko, E., et al. (2009). Diffusion-tensor imagingimplicates prefrontal axonal injury in executive function impair-ment following very mild traumatic brain injury. Radiology, 252(3), 816–824.
Lishman, W. A. (1988). Physiogenesis and psychogenesis in the 'post-concussional syndrome'. The British Journal of Psychiatry, 153,460–469.
Little, D. M., Kraus, M. F., et al. (2010). Thalamic integrity underliesexecutive dysfunction in traumatic brain injury. Neurology, 74(7),558–564.
Liu, A. Y., Maldjian, A. J., et al. (1999). Traumatic brain injury:diffusion-weighted MR imaging findings. AJNR. American Jour-nal of Neuroradiology, 20(9), 1636–1641.
Liu, M. C., Akinyi, L., Scharf, D., et al. (2010). Ubiquitin C-terminal hydroslase-L1 as a biomarker for ischemic and trau-matic brain injury in rats. European Journal of Neuroscience,31, 722–732.
Ljungqvist, J., Nilssonn, D., et al. (2011). Longitudinal study of thediffusion tensor imaging properties of the corpus callosum inacute and chronic diffuse axonal injury. Brain Injury, 25(4),370–378.
Lo, C., Shifteh, K., et al. (2009). Diffusion tensor imaging abnormal-ities in patients with mild traumatic brain injury and neurocogni-tive impairment. Journal of Computer Assisted Tomography, 33(2), 293–297.
MacDonald, C. L., Johnson, A. M., et al. (2011). Detection of blast-related traumatic brain injury in U. S. military personnel. The NewEngland Journal of Medicine, 364(22), 2091–2100.
Machulda, M. M., Bergquist, T. F., et al. (1988). Relationshipbetween stress, coping, and postconcussion in a healthy adultpopulation. Archives of Clinical Neuropsychology, 13(5),415–424.
MacKenzie, J. D., Siddiqi, F., et al. (2002). Brain atrophy in mild ormoderate traumatic brain injury: a longitudinal quantitative anal-ysis. AJNR. American Journal of Neuroradiology, 23(9), 1509–1515.
Malcolm, J. G., Shenton, M. E., & Rathi, Y. (2010). Filtered multi-tensor tractography. IEEE Transactions on Medical Imaging, 29,1664–1675.
Brain Imaging and Behavior (2012) 6:137–192 189
Maller, J. J., Thomson, R. H., et al. (2010). Traumatic brain injury,major depression, and diffusion tensor imaging: making connec-tions. Brain Research Reviews, 64(1), 213–240.
Maruta, J., Suh, M., et al. (2010). Visual tracking synchronization as ametric for concussion screening. The Journal of Head TraumaRehabilitation, 25(4), 293–305.
Matsushita, M., Hosoda, K., et al. (2011). Utility of diffusion tensorimaging in the acute stage of mild to moderate traumatic braininjury for detecting white matter lesions and predicting long-termcognitive function in adults. Journal of Neurosurgery, 115(1),130–139.
Matthews, S. C., Strigo, I. A., et al. (2011). A multimodal imagingstudy in U. S. veterans of Operations Iraqi and Enduring Freedomwith and without major depression after blast-related concussion.NeuroImage, 54(Suppl 1), S69–S75.
Mayer, A. R., Ling, J., et al. (2010). A prospective diffusion tensorimaging study in mild traumatic brain injury. Neurology, 74(8),643–650.
McAllister, T. W., Ford, J. C., et al. (2012). Maximum principal strainand strain rate associated with concussion diagnosis correlateswith changes in corpus callosum white matter indices. Annals ofBiomedical Engineering, 40(1), 127–140.
Messe, A., Caplain, S., et al. (2011). Diffusion tensor imaging andwhite matter lesions at the subacute stage in mild traumatic braininjury with persistent neurobehavioral impairment. Human BrainMapping, 32(6), 999–1011.
Metzler-Baddeley, C., O'Sullivan, M. J., Bells, S., et al. (2012). Howand how not to correct for CSF-contamination in diffusion MRI.NeuroImage, 59, 1394–1403.
Miles, L., Grossman, R. I., et al. (2008). Short-term DTI predictors ofcognitive dysfunction in mild traumatic brain injury. Brain Injury,22(2), 115–122.
Miller, L. (1996). Neuropsychology and pathophysiology of mildhead injury and the postconcussive syndrome: clinical andforensic considerations. Journal of Cognitive Rehabilitation,14, 8–23.
Mitra, P. P. (1992). Diffusion propagator as a probe of the structure ofporous media. Physical Review Letters, 68(24), 3555–3558.
Mittl, R. L., Garossman, R. I., et al. (1994). Prevalence of MR evi-dence of diffuse axonal injury in patients with mild head injuryand normal head CT findings. AJNR. American Journal of Neu-roradiology, 15(8), 1583–1589.
Mondello, S., Muller, U., Jeromin, A., et al. (2011). Blood-baseddiagnostics of traumatic brain injuries. Expert Review of Molecu-lar Diagnostics, 11(1), 65–78.
Mori, S., Crain, B. J., Chacko, V. P., et al. (1999). Three dimensionaltracking of axonal projections in the brain by magnetic resonanceimaging. Annals of Neurology, 45, 265–269.
Nelles, M., Block, W., et al. (2008). Combined 3T diffusion tensortractography and 1H-MR spectroscopy in motor neuron disease.American Journal of Neuroradiology, 29(9), 1708–1714.doi:10.3174/ajnr.A1201.
Niogi, S. N., & Mukherjee, P. (2010). Diffusion tensor imaging ofmild TBI. The Journal of Head Trauma Rehabilitation, 25(4), 241–255.
Niogi, S. N., Mukherjee, P., et al. (2008a). Extent of microstructuralwhite matter injury in postconcussive syndrome correlates withimpaired cognitive reaction time: a 3T diffusion tensor imagingstudy of mild traumatic brain injury. AJNR. American Journal ofNeuroradiology, 29(5), 967–973.
Niogi, S. N., Mukherjee, P., et al. (2008b). Structural dissociation ofattentional control and memory in adults with and without mildtraumatic brain injury. Brain, 131(12), 3209–3221.
Nolin, P., & Heroux, L. (2006). Relations among sociodemo-graphic, neurologic, clinical, and neuropsychologic variables,and vocational status following mild TBI: a follow-up study.
The Journal of Head Trauma Rehabilitation, 21(6), 514–526.
Okie, S. (2005). Traumatic brain injury in the war zone. New EnglandJournal of Medicine, 352, 2043–2047.
Oppenheimer, D. R. (1968). Microscopic lesions in the brain followinghead injury. Journal of Neurology, Neurosurgery & Psychiatry,31, 299–306.
Ozarian, M. E. (2003). Generalized diffusion tensor imaging andanalytical relationships between diffusion tensor imaging andhigh angular resolution diffusion imaging. Magnetic Resonancein Medicine, 50(5), 955–965.
Park, J. H., Park, S. W., et al. (2009). Detection of traumatic cerebralmicrobleeds by susceptibility-weighted image of MRI. Journal ofKorean Neurosurgical Society, 46, 365–369.
Pasternak, O., Assaf, Y., et al. (2008). Variational multiple-tensorfitting of fiber-ambiguous diffusion-weighted magnetic reso-nance imaging voxels. Magnetic Resonance Imaging, 26(8),1133–1144.
Pasternak, O., Sochen, N., et al. (2009). Free water elimination andmapping from diffusion MRI. Magnetic Resonance in Medicine,62(3), 717–730.
Pasternak, O., & Bouix, S., et al. (2010). Diffusion imaging revealstwo spatially separable mechanisms. In T. B. I. Mild (Ed.), The3rd federal interagency conference on TBI. Washington, DC.
Pasternak, O., & Kubicki, O., et al. (2011a). Identification of neuro-inflammation in mild traumatic brain injuries using a free-wateratlas. Annual Meeting of the Organization for the Human BrainMapping.
Pasternak, O., Westin, C.-F., et al. (2011). Free water modulation ofwhite matter integrity measures - with application to schizophre-nia. Proceedings International Society of Magnetic Resonance inMedicine, 19, 2544.
Pfefferbaum, A., Sullivan, E. V., et al. (2000). In vivo detection andfunctional correlates of white matter microstructural disruption inchronic alcoholism. Alcoholism, Clinical and Experimental Re-search, 24(8), 1214–1221.
Pierpaoli, C., & Basser, P. J. (1996). Toward a quantitative assessmentof diffusion anisotropy. Magnetic Resonance in Medicine, 36,893–906.
Pohl, K. M., Bouix, S., et al. (2007). A hierarchical algorithm for MRbrain image parcellation. IEEE Transactions on Medical Imaging,26(9), 1201–1212.
Povlishock, J. T., & Coburn, T. H. (1989). Morphopathological changeassociated with mild head injury. In H. S. Levin, H. M. Eisenberg,& A. L. Benton (Eds.), Mild head injury (pp. 37–52). New York:Oxford University.
Provencher, S. W. (2001). Automatic quantitation of localized invivo1H spectra with LCModel. NMR in Biomedicine, 14(4),260–264.
Reichenbach, J. R., Markus, B., et al. (2000). High-resolution MRvenography at 3.0 tesla. Journal of Computer Assisted Tomogra-phy, 24(6), 949–957.
Rimel, R. W., Giordani, B., et al. (1981). Disability caused by minorhead injury. Neurosurgery, 9(3), 221–228.
Ross, B. D., Ernst, T., et al. (1998). 1H MRS in acute traumaticbrain injury. Journal of Magnetic Resonance Imaging, 8(4),829–840.
Ross, B., et al. (2005). MR spectroscopy of hypoxic brain injury. In J.Gillard, A. Waldman, & P. B. Barker (Eds.), Clinical MR neuro-imaging: Diffusion, perfusion and spectroscopy (pp. 690–705).Cambridge: Cambridge University Press.
Ruff, R. M., Camenzuli, L., et al. (1996). Miserable minority: emo-tional risk factors that influence the outcome of a mild TBI. BrainInjury, 10(8), 551–565.
Rutgers, D. R., Fillard, P., et al. (2008). Diffusion tensor imagingcharacteristics of the corpus callosum in mild, moderate, and
190 Brain Imaging and Behavior (2012) 6:137–192
severe traumatic brain injury. AJNR. American Journal of Neuro-radiology, 29(9), 1730–1735.
Rutgers, D. R., Toulgoat, F., et al. (2008). White matter abnor-malities in mild traumatic brain injury: a diffusion tensorimaging study. AJNR. American Journal of Neuroradiology,29(3), 514–519.
Salmond, C. H., Menon, D. K., et al. (2006). Diffusion tensor imagingin chronic head injury survivors: correlations with learning andmemory indices. NeuroImage, 29(1), 117–124.
Scheid, R., Preul, C., et al. (2003). Diffuse axonal injury associatedwith chronic traumatic brain injury, evidence from T2*-weightedgradient-echo imaging at 3T. AJNR. American Journal of Neuro-radiology, 24, 1049–1056.
Scheid, R., Ott, D. V., et al. (2007). Comparative magnetic resonanceimaging at 1.5 and 3 Tesla for the evaluation of traumatic micro-bleeds. Journal of Neurotrauma, 24(12), 1811–1816.
Schonberger, M., Ponsford, J., et al. (2009). The Relationship betweenage, injury severity, and MRI findings after traumatic brain injury.Journal of Neurotrauma, 26, 2157–2167.
Seeger, U., Klose, U., et al. (2003). Parameterized evaluation of macro-molecules and lipids in proton MR spectroscopy of brain diseases.Magnetic Resonance in Medicine, 49(1), 19–28.
Shenton, M. E., Whitford, T. J., et al. (2010). Structural neuroimagingin schizophrenia: from methods to insights to treatments. Dia-logues in Clinical Neuroscience, 12(3), 269–332.
Shutter, L., Tong, K. A., et al. (2004). Proton MRS in acute TBI: rolefor glutamate/glutamine and choline for outcome prediction.Journal of Neurotrauma, 21(12), 1693–1705.
Singh, M., Jeong, J., et al. (2010). Novel diffusion tensor imagingmethodology to detect and quantify injured regions and affectedbrain pathways in traumatic brain injury. Magnetic ResonanceImaging, 28(1), 22–40.
Smith, D. H., Meaney, D. F., et al. (1995). New magnetic resonanceimaging techniques for the evaluation of traumatic brain injury.Journal of Neurotrauma, 12(4), 573–577.
Smith, S. M., Jenkinson, M., et al. (2006). Tract-based spatial statistics:voxelwise analysis of multi-subject diffusion data. NeuroImage,31(4), 1487–1505.
Smits, M., Houston, G. C., et al. (2011). Microstructural brain injury inpost-concussion syndrome after minor head injury. Neuroradiol-ogy, 53(8), 553–563.
Song, S. K., Sun, S. W., et al. (2001). Dysmyelination revealed throughMRI as increased radial (but unchanged axial) diffusion of water.NeuroImage, 17(3), 1429–1436.
Song, S. K., Sun, S. W., et al. (2003). Diffusion tensor imaging detectsand differentiates axon and myelin degeneration in mouse opticnerve after retinal ischemia. NeuroImage, 20(3), 1714–1722.
Sosin, D. M., Sniejek, J. E., et al. (1996). Incidence of mild andmoderate brain injury in the United States. Brain Injury,10, 47–54.
Sponheim, S. R., McGuire, K. A., et al. (2011). Evidence of disruptedfunctional connectivity in the brain after combat-related blastinjury. NeuroImage, 54(Suppl 1), S21–S29.
Stein, M. B., & McAllister, T. W. (2009). Exploring the convergence ofposttraumatic stress disorder and mild TBI. The American Journalof Psychiatry, 166(7), 768–776.
Stern, R. A., & Riley, D. A., et al. (2011). Long-term consequences ofrepetitive brain trauma: Chronic traumatic encephalopathy. PM R.
Strangman, G. E., O'Neil-Pirozzi, T. M., et al. (2010). Regionalbrain morphometry predicts memory rehabilitation outcomeafter traumatic brain injury. Frontiers in Human Neuroscience,4, 182.
Tang, C. Y., Friedman, et al. (2007). Correlations between DiffusionTensor Imaging (DTI) and Magnetic Resonance Spectroscopy(1H MRS) in schizophrenic patients and normal controls. BMCPsychiatry, 7(1), 25. doi:10.1186/1471-244X-7-25.
Tanielian, T., & Jaycox, L. H. (2008). Invisible wounds of war:Psychological and cognitive injuries, their consequences andservices to assist recovery. Santa Monica: CAO: RANDCorp.
Tate, D. F., & Bigler, E. D. (2000). Fornix and hippocampalatrophy in traumatic brain injury. Learning and Memory, 7(6), 442–446.
Teasdale, G., & Jennett, B. (1974). Assessment of coma and impairedconsciousness. A practical scale. Lancet, 2(7872), 81–84.
Thurman, D. J. (2001). The epidemiology and economics of headtrauma. In L. Miller & R. Hayes (Eds.), Head trauma: Basic,preclinical, and clinical directions (pp. 324–347). NY: Wiley.
Tollard, E., Galanaud, D., Perlbarg, V., et al. (2009). Experience ofdiffusion tensor imaging and 1H spectroscopy for outcomeprediction in severe traumatic brain injury: preliminary results.Critical Care Medicine, 37(4), 1448–1455.
Tong, K. A., Ashwal, S., et al. (2003). Hemorrhagic shearing lesions inchildren and adolescents with posttraumatic diffuse axonal injury:improved detection and initial results. Radiology, 227(2), 332–339.
Tong, K. A., Ashwal, S., et al. (2004). Diffuse axonal injury inchildren: clinical correlation with hemorrhagic lesions. Annals ofNeurology, 56(1), 36–50.
Tournier, J. D., Yeh, C. H., Calamante, F., et al. (2008). Resolvingcrossing fibres using constrained spherical deconvolution: valida-tion using diffusion-weighted imaging phantom data. Neuro-Image, 42(2), 617–624.
Trivedi, M. A., Ward, M. A., et al. (2007). Longitudinal changes inglobal brain volume between 79 and 409 days after traumaticbrain injury: relationship with duration of coma. Journal ofNeurotrauma, 24, 766–771.
Tuch, D. S., Reese, T. G., et al. (2002). High angular resolutiondiffusion imaging reveals intravoxel white matter fiber heteroge-neity. Magnetic Resonance in Medicine, 48(4), 577–582.
Vagnozzi, R., Signoretti, S., et al. (2010). Assessment of metabolicbrain damage and recovery following mild TBI: a multicentre,proton magnetic resonance spectroscopic study in concussedpatients. Brain, 133(11), 3232–3242.
Vanderploeg, R. D., Curtiss, G., et al. (2007). Long-term morbiditiesfollowing self-reported mild traumatic brain injury. Journal ofClinical and Experimental Neuropsychology, 29, 585–598.
Warden, D. (2006). Military TBI during Iraq and Afghanistan wars.The Journal of Head Trauma Rehabilitation, 21, 398–402.
Warner, M. A., Marquez de la Plata, C., et al. (2010). Assessing spatialrelationships between axonal integrity, regional brain volumes,and neuropsychological outcomes after traumatic axonal injury.Journal of Neurotrauma, 27(12), 2121–2130.
Warner, M. A., Youn, T. S., et al. (2010). Regionally selective atrophyafter traumatic axonal injury. Archives of Neurology, 67(11),1336–1344.
Westin, C.-F., Maier, S. E., Khidhir, B., Everett, P., Jolesz, F. A., &Kikinis, R. (1999). Image processing for diffusion tensor magneticresonance imaging. In Medical Image Computing and Computer-
Whitford, T. J., & Kubicki, M., et al. (2011). Diffusion tensor imaging,structural connectivity, and schizophrenia. Schizophr Res Treatment,1–7.
Wilde, E. A., Bigler, E. D., et al. (2004). Alcohol abuse and traumaticbrain injury: quantitative magnetic resonance imaging and neuro-psychological outcome. Journal of Neurotrauma, 21, 137–147.
Wilde, E. A., Bigler, E. D., et al. (2006). Post-traumatic amnesiapredicts long- term cerebral atrophy in traumatic brain injury.Brain Injury, 20, 695–699.
Yount, R., Raschke, K. A., et al. (2002). Traumatic brain injury andatrophy of the cingulate gyrus. The Journal of Neuropsychiatryand Clinical Neurosciences, 14, 416–442.
Yu-Chien, W. U., Andrew, L., et al. (2007). Hybrid diffusion imaging.NeuroImage, 36(3), 617–629.
Brain Imaging and Behavior (2012) 6:137–192 191
Yurgelun-Todd, D. A., Bueler, E., et al. (2011). Neuroimaging corre-lates of traumatic brain injury and suicidal behavior. The Journalof Head Trauma Rehabilitation, 26(4), 276–289.
Zhang, K., Johnson, B., et al. (2010). Are functional deficits in con-cussed individuals consistent with white matter structural alter-ations: combined FMRI & DTI study. Experimental BrainResearch, 204(1), 57–70.
Zhou, J., Xu, S., Proctor, J., et al. (2012). Diffusion Kurtosis as an invivo imaging marker for reactive astrogliosis in traumatic braininjury. NeuroImage, 59, 467–477.
Zou, K. H., Greve, D. N., et al. (2005). Reproducibility of functionalMR imaging: preliminary results of prospective multi-institutionalstudy performed by biomedical informatics research network.Radiology, 237(3), 781–789.
192 Brain Imaging and Behavior (2012) 6:137–192