12
Structural and Functional Brain Imaging in Mild Traumatic Brain Injury
Jeffrey David Lewine
BACKGROUND
• Almost 85% of brain injuries are classified as mild. However, such classification does not mean that the injury is insignificant.
• Relevant imaging strategies include both structural and functional methods, with specific recommendations varying as the patient progresses through the acute (first 24–96 hours), subacute (up to 3 months), and chronic (beyond 3 months) phases of injury.
• Anatomic neuroimaging examinations using computed tomography (CT) and/or magnetic resonance imaging (MRI) are usually negative in cases of mild traumatic brain injury (MTBI). When CT or MRI are positive for structural pathology in the acute setting of MTBI, the injury is classified as a complicated MTBI.
• Advanced structural and functional brain imaging methods can sometimes provide objective evidence of MTBI in the absence of findings on conventional imaging.
IMAGING ASSESSMENTS
Types of Imaging
There are two main types of imaging methods—those that examine brain structure (CT and MRI), and those that look at brain function. Functional imaging methods include those that examine biochemical, metabolic, and hemodynamic functions, and those that directly assess brain electrophysiology. The main biochemical/metabolic/hemodynamic methods are: positron emission tomography (PET), single photon emission computed tomography (SPECT), functional MRI (fMRI), and magnetic resonance spectroscopy (MRS). The main electrophysiological methods are electro-encephalography (EEG) and magnetoencephalography (MEG) [1]. Functional imaging is not typically warranted in routine clinical management of typical patients with MTBI, although it may be of utility for clinically complicated/atypical cases (e.g., those with seizures or those in which other diagnoses are also being entertained), medicolegal purposes, and/or in the context of a specific research question.
Structural Imaging
• CT—CT uses x-rays to make tomographic images of the body. CT is generally the first brain-imaging method to be employed in the emergency medical management of patients with head trauma, because it provides rapid identification of skull fractures and intracranial bleeds [2,3].
Findings in MTBI: CT examination is generally within normal limits in cases of MTBI, although intracranial bleeds are sometimes seen. Nevertheless, CT is almost always recommended in cases of head trauma with loss of consciousness, to rule out potentially life-threatening complications.
• MRI—By examining how systematically applied magnetic field gradients and radiofrequency pulses alter the behavior of the hydrogen protons of water molecules, MRI provides detailed information on the soft tissues of the body. MRI is inferior to CT for identification of skull fractures and acute bleeding, but it is superior to CT for the identification of intraparenchymal abnormalities, including diffuse axonal and shear injuries.
Findings in MTBI: Most clinical MRI evaluations in MTBI are within normal limits, but sometimes clinically meaningful abnormalities are identified. Abnormal findings in MTBI may include identification of hematomas, contusions, and FLAIR/T2 white matter hyperintensities near the gray-white matter junction, suggestive of shear injury. In the chronic phase of injury, quantitative MRI may reveal evidence of generalized and/or regional atrophy/volume reductions, even in MTBI [4–8]. When available, two additional MRI protocols may be considered. These are diffusion-weighted imaging [DWI, most commonly in the form of diffusion tensor imaging (DTI)], and recently developed susceptibility-weighted imaging (SWI) [9–12]. DWI can provide important information on diffuse axonal injury and the integrity of white matter pathways [as indexed by fractional anisotropy (FA) values], while SWI allows for identification of microbleeds that are not readily seen using standard imaging sequences (see Figure 12.1). Unfortunately, these protocols require software and hardware that is often not available at community-based hospitals.
FIGURE 12.1 3T magnetic resonance images of the brain of a female patient (38 years old) 2 months after a traumatic brain injury. (A) T2 flair imaging shows no abnormalities. (B) Gradient echo imaging shows three lesions in the right frontal lobe (white arrows in white circle). (C) Susceptibility-weighted imaging shows more low intensity lesions at the same location and also in the corpus callosum (white circle). (D) Susceptibility-weighted imaging mapping shows high-intensity lesions (white circle) and deep veins (white arrow). Source: From Ref. [11]. Liu J, Kou Z, Tian Y, Diffuse axonal injury after traumatic cerebral microbleeds: an evaluation of imaging techniques, Neural Regen Res. 2014;9(12):1222–1230. With permission of Neural Regeneration Research.
Functional Imaging
• PET—PET uses compounds labeled with positron-emitting radionucleotides to assess brain biochemistry and metabolism. Brain injury evaluations usually use fluorodeoxyglucose (FDG), and thereby measure regional metabolism, an indirect measure of neuronal activity. Clinical studies typically collect data while the brain is “at rest,” whereas research and medicolegal studies often collect additional data during the performance of working memory or attention tasks.
Findings in MTBI: Even in cases without gross structural damage, PET images are often reported to be abnormal for days to months post MTBI. Regions of hypometabolism are most commonly identified, but both hypo- and hyper-metabolic regions may be reported, with a variable correlation between specific PET findings and each patient’s neuropsychological profile [13–16]. The role of PET in clinical management of MTBI is yet to be determined.
• SPECT—SPECT is similar to PET except that relevant radionucleotides emit single photons (gamma rays) rather than positrons. As such, the spatial resolution of SPECT is slightly poorer than that of PET. On the other hand, the dose of radioactivity is lower for SPECT, so it is possible to perform resting and activation studies on successive days, with intrasubject comparisons. Most trauma studies use technitium-labeled hexamethylpropyleneamine oxime (Tc-HMPAO), with images providing information on regional blood flow, an indirect measure of regional metabolism.
Findings in MTBI: Even in cases without gross structural damage on MRI, visual inspection of SPECT data leads to reports of abnormalities in the majority of cases of TBI when all injury severity levels are pooled in aggregate, with quantitative methods showing moderate sensitivity to MTBI [17–19]. Frontal and temporal hypoperfusion, along with basal ganglia hypoperfusion are most commonly reported [17], but the scientific data are conflicted with respect to the consistency of the relationships between the location of SPECT anomalies and specific clinical profiles.
• fMRI—The magnetic properties of oxygenated and deoxygenated blood are slightly different, so it is possible to use echo-planar, blood oxygen-level dependent (BOLD) MRI to examine regional blood flow. By comparison of data collected during rest versus active states, information can be obtained concerning how the brain activates during specific sensory, motor, or cognitive tasks. Evaluations in MTBI typically use working memory or attention tasks. There is also emerging work indicating MTBI-associated abnormalities in resting state connectivity profiles.
Findings in MTBI: Several investigative teams have shown, in group-averaged data, atypical activation patterns during cognitive challenges related to working memory or attention. Observations include MTBI-related reductions in activity in specific network nodes, or recruitment of brain regions not normally associated with completion of a particular task [20–22]. Data are also emerging to show altered patterns of functional connectivity at rest [23–25]. Generalization from group data to an individual patient is problematic. Diagnostic applicability of fMRI is generally undetermined. At the moment, fMRI is a valuable research method, but its utility in the clinical management of individual patients with MTBI remains to be demonstrated.
• MRS—The local microenvironment influences the resonance frequency of protons, so it is possible to use magnetic resonance (MR) technology to measure regional concentrations of certain metabolites including N-acetylaspartate (NAA, a neuron-specific metabolic marker), creatine (Cr, which is related to energy metabolism), and choline (Cho, a cell membrane marker).
Findings in MTBI: Some studies have identified altered metabolite concentrations and reduced NAA/Cho and NAA/Cr ratios following mild trauma, suggesting perturbed metabolic activity in the regions measured [26–28]. Clinical applicability is still under investigation.
• EEG—EEG uses contact electrodes applied to the scalp to measure electrical potential patterns that are directly caused by changing patterns of current within brain cells. EEG data may be visually inspected for epileptiform transients and focal or diffuse slow waves, and it may be subjected to quantitative analyses (qEEG) with comparison to a normative database. From a clinical management perspective for MTBI, EEG is only warranted in those cases where posttraumatic epilepsy is suspected.
Findings in MTBI: Visual inspection of EEG following MTBI reveals abnormalities in less than 25% of cases, and these usually consist of nonspecific diffuse slowing of uncertain clinical significance. qEEG may be abnormal in cases of MTBI [29–34], but there are often concerns about the etiologic specificity of such findings with respect to other clinical conditions including depression and substance use disorders [33,34].
• MEG—MEG uses special superconducting sensors to detect the weak neuromagnetic signals generated by the brain’s electrical activity [35]. Simultaneous EEG data can be collected to take advantage of the complimentary nature of the methods. Both spontaneous and stimulus evoked activity may be gathered. Normally functioning awake brain tissue mostly generates oscillating electromagnetic signals in the theta-gamma range (6–60 Hz), with little power in the 1–4 Hz delta range. Therefore, the presence of delta-band, focal dipolar slow wave activity (DSWA) is a reliable sign of dysfunctional tissue [35–37].
Findings in MTBI: In contrast to normal controls, DSWA is commonly seen in tissue that appears normal on MRI, especially if there are persistent cognitive symptoms [38–43] (see Figure 12.2). The presence of DSWA tracks with symptom severity and persistence, and it appears in a pattern consistent with known neuroanatomical–cognitive relationships [40,42]. MEG slow waves resolve when symptoms resolve, and persist for even years after injury when symptoms persist [38]. MEG slow waves appear to be an objective correlate of MTBI-perturbed brain function, but in considering this, it must be recognized that other conditions (stroke, epilepsy, etc.) can also produce focal slowing [35–37]. The presence of these other conditions must therefore be ruled out before an observation of slow waves can be construed to mean that a TBI has occurred, or that it is responsible for the abnormalities detected. As is the case with the other functional modalities discussed earlier, the role of MEG in the clinical management of patients with MTBI is still under investigation.
FIGURE 12.2 Use of MEG in the evaluation of mild traumatic brain injury (MTBI). (A) For this patient with a history of mild head trauma and persistent postconcussive symptoms (PCS), large amplitude slow waves are seen. The neuronal generators of the slow wave activity can be localized and plotted on spatially aligned magnetic resonance (MR) images. (B) Data are shown for a 23-year-old female subject with a MTBI who reported persistent memory and attention problems. Magnetic resonance imaging (MRI) was within normal limits. Single photon emission computed tomography (SPECT) revealed mild bitemporal hypoperfusion. Magnetoencephalography (MEG) at 21 months revealed right and left temporal focal slow waves, right temporal spikes, and additional left parietal slowing, all in a pattern consistent with documented neuropsychological deficits. Source: Adapted from Ref. [40]. Lewine JD, Davis JT, Bigler E, et al. Objective documentation of traumatic brain injury subsequent to mild head trauma: multimodal brain imaging with MEG, SPECT, and MRI. J Head Trauma Rehabil. 2007;23:141–155. Editor’s note: Caution should be exercised in assuming that these findings are solely due to an isolated MTBI, or that they can be generalized to all MTBI patients (see earlier discussion).
TYPE/TIMING OF IMAGING
The Acute Phase of Injury—Structural Imaging
In the acute care setting, well-established clinical criteria exist (New Orleans Head CT Criteria, Canadian Head CT Rule) to aid in determining which patients may require CT imaging [2,3]. (Specific clinical guidelines delineating indications for imaging in mild TBI are addressed further in Chapter 7.) According to most published guidelines, acute MRI is generally warranted only if there is a significant deterioration in clinical status [2,3]. The one exception comes from the Defense and Veterans Brain Injury Center (DVBIC), which recommends MRI if there is a memory loss of greater than 15 minutes and persistent symptoms for more than 72 hours, or a history of multiple prior concussions during the past 12 months (see electronic references).
Subacute Phase of Injury—Structural Imaging
At present, there are no specific published guidelines with respect to recommendations for imaging in the subacute phase. There appears to be an unofficial consensus that MRI (or repeat CT if MRI is not available) is warranted if there is any type of neurological deterioration during the subacute phase, or there are persistent neurological deficits beyond 2 to 3 months that are not well explained by clinical context and initial imaging.
Chronic Phase of Injury—Structural and Functional Imaging
The evaluation and treatment of MTBI patients who have symptoms that persist into the chronic period (>3 months) is especially challenging. There is often a disconnection between the self-reported severity of symptoms on subjective questionnaires versus what can be documented using objective and formal neurological and/or neuropsychological evaluation.
Standard CT and MR structural imaging are typically within normal limits in the chronic period, even for patients diagnosed as having had a complicated MTBI. On the other hand, advanced structural methods do show some continuing sensitivity to injury during the chronic period.
Hemosiderin (iron) deposits associated with microbleeds into the brain parenchyma are not readily reabsorbed, so old bleeds can be identified using SWI, even decades after the initial bleed. However, this means that caution must be used in trying to link SWI findings to a particular remote traumatic event. The data must be carefully considered within a framework of the patient’s entire medical history.
Emerging data suggest that DWI strategies can be informative in evaluating the chronic phase of MTBI, with group evidence of reduced FA—especially in the corpus callosum [9,10], but application of this approach to individual subjects requires the use of quantitative statistical methods. Also, it is important to note that many non-TBI–related factors can impact FA values [44,45].
Quantitative MRI approaches in which brain or gray matter volumes from an individual patient are evaluated statistically with respect to normative data sets are also starting to show promise [4–8].