Approaches for Monitoring Warfighter Blast-related Exposures in Training to Develop Effective Safety Standards

ABSTRACT

Introduction

Traumatic brain injuries are of concern to the sports and military communities because of the age of the participants and costly burden to society. To markedly reduce the impact of traumatic brain injury and its sequela (TBI-S), it is necessary to determine the initial vulnerability of individuals as well as identify new technologies that indicate early signs of TBI-S.

Materials and Methods

Currently, diverse methods have been used by the authors and others in laboratory settings to reveal early signs of persistent TBI-S including simulation modeling of the effect of rapid deceleration on the deviatoric strain (shear force) imposed on specific brain regions, auditory evoked potential (AEP) measurements to determine injury to the auditory cortex optokinetic nystagmus (OKN) measures sensitive to vestibular trauma, and optical coherence tomography (OCT) measures that reveal changes in central visual function obtained noninvasively by examination of the retina.

Results

Simulation studies provided technical information on maximal deviatoric strain at the base of the sulci and interface of gray and white matter consistent with results from neuropathology and from magnetic resonance imaging. The AEP and OKN reveal measurable injury to similar regions below the Sylvian fissure including auditory cortex and midbrain, and the OCT reveals changes to the retina consistent with forceful deceleration effects.

Conclusions

The studies and results are consistent with prior work demonstrating that noninvasive tests may be sensitive to the presence of TBI-S, potentially in the training field as advances in the portability of test instruments are underway. When combined with baseline data gathered from individuals in quantitative form, key variances can emerge. Therefore, it is hypothesized that AEP, OKN, and OCT, taken together, may yield faster objective and quantitative neurophysiological measures serving as a “signature” of neural injury and more indicative of potentially persistent TBI-S—recommending larger scale longitudinal studies.

INTRODUCTION

Traumatic brain injury (TBI) and persistent traumatic brain injury and its sequela (TBI-S) are of major concern to the military communities because of the age of the service members (18-40 years) and costly impact to force response. Younger age of onset of TBI-S may result in reduced ability to continue to serve in desired roles and a longer duration of symptoms and treatment including disability. The Centers for Disease Control and Prevention estimate that 1.7 million persons in the United States per year experience TBI.¹ The Department of Defense reports that between the years 2000 and 2018 approximately 25,000 service members experienced TBI each year (Defense Medical Surveillance System, Defense and Veterans Brain Injury Center, 2018). Although the majority of individuals experiencing repetitive or even a single severe blast TBI will recover and lead a full and productive life, approximately 30% of affected individuals will progress to chronic TBI-S and in these cases dementia may be experienced within 15 to 20 years.

At the present time, technology does not exist to differentiate between transient episodes of TBI and those TBI events that precede the development of persistent TBI-S. Prior studies from many laboratories, including our own, have shown that TBI resulting from blast and overpressure have direct and profound effects upon two primary sensory systems in the brain: visual and auditory pathways of the central nervous system.^2–5

The marked increase in incidence of TBI in athletes engaged in contact sports and in service members returning from deployment over the past two decades has led to extensive research regarding the mechanism causing this injury and potential treatments for the condition.^6–12 In the affected population, inflammatory changes in the brain including activation of microglia, generation of antibodies to neuronal and glial proteins, increased permeability of the blood brain barrier, and changes in the expression of interleukins and cytokines are associated with the clinical status of the patient.^11,13 Neuropathological studies on brain samples recovered from patients who died following diagnosis with severe TBI revealed changes in the brain parenchyma that are designated as chronic traumatic encephalopathy (CTE).¹⁰

Although research into CTE continues with more longitudinal focus through the National Institutes of Health, there continues to be a need for additional research on CTE discovered between athletes and nonathletes—service members and civilians—with a search for factors like TBI-S and functional predictive tests in the fields of audiology and optical responses. Chronic traumatic encephalopathy is currently a diagnosis that is obtained postmortem and is dependent upon brain tissue samples that are examined microscopically. McKee and colleagues define CTE as a “tauopathy, characterized by the deposition of hyperphosphorylated tau protein as neurofibrillary tangles, astrocytic tangles and neurites in striking clusters around small blood vessels of the cortex, typically at the sulcal depths.”¹⁰

Of primary interest to the current hypothesis is the observation that the lesions in CTE cluster around small vessels of the cortex localized at the sulcal depths rather than at the apices of the gyri. During rapid deceleration of the head associated with forceful impacts, the apices of the gyri are the sites of initial contact with the calvarium and these regions might be anticipated to be the major site of injury; however, this is not necessarily the case. In an earlier report, Kornguth et al.¹¹ proposed that the depths of the sulci will likely be the primary site of injury from traumatic head impacts as a result of a “water hammer” effect following the forceful impact of the brain upon a rigid bone (calvarium) and the incompressibility of the cerebral spinal fluid (water) that is driven into the sulci where the energy is dissipated at the base of the sulcus. The resulting forces at the base of the sulci were hypothesized to cause shearing at the base, thereby causing tears in the regional microvessels with microbleeds as a consequence.

Fagan et al. in research sponsored by the U.S. Army Combat Capabilities Development Command Army Research Laboratory¹⁴ undertook a computational modeling study to investigate the load transmission to the brain under blunt impact conditions in order to understand the load amplification mechanism at the base of the sulcus, which was also observed in Kornguth et al.¹¹ In this study, the effects of brain geometry on impact-induced stress propagation, especially at the base of the sulci, were examined and compared with the origin of structural changes in this region as reported by Kornguth et al.¹¹ The 2D finite element models of a human head were developed and designed to simulate impact between a human head and a rigid wall at a prescribed initial velocity with various brain-structure approximations: smooth brain, brain with sulci and gyri with and without white and gray matter differentiation. The Activation Likelihood Estimate method was used to model the cerebrospinal fluid (CSF) as a deforming fluid while treating the skull and the brain as deforming solids using Lagrangian discretization. This method is better suited to model the fluid–structure interaction between the CSF and the adjacent skull and brain in comparison with a purely Lagrangian method. The displacement, stress, and strain in the head arising from the impact event were computed.

These simulations showed elevated levels of deviatoric strain and first and third principal strains near the bases of the sulci compared with the other areas in the brain (Fig. 1). These regions of intense maximum deviatoric strain near the base of the sulci compare well to the regions of microbleeds observed during the Kornguth et al.¹¹ study focused on female collegiate soccer players. Even though CSF was not observed to be driven into the sulci creating higher pressure, elevated maximum deviatoric strain levels near the base of the sulci were exhibited. This shearing of tissue close to the base of the sulci especially at the interface of white and gray brain matter, and the location of these regions of elevated levels of strain qualitatively matched the microvessel damage and microbleed patterns clinically observed in Kornguth et al.¹¹ study. Additional differentiation between white and gray matter into the model led to a larger area of the brain that experienced a greater maximum deviatoric strain.

FIGURE 1.

Typical strain amplification observed at the bases of sulci in two models. (Left) Model with homogeneous brain. (Right) Model with white and gray matter differentiation.

Whether the “water hammer” effect or the differing inertial properties of the gray and white matter following rapid deceleration is the dominant factor causing lesions at the base of the sulci, the most affected region will be near the base of the deepest sulcus, the Sylvian fissure. The auditory evoked potential (AEP) and optokinetic nystagmus (OKN) are functional measures that will reveal changes in brain integrity near the base of the Sylvian fissure, and for this reason, we propose the combination of these measures to assess the TBI-S population.

Auditory complaints are commonly expressed by service members who experienced TBI and may, in part, be because of the water hammer effect described in Kornguth et al.¹¹ These deficits may occur even in the presence of peripheral normal audiometric hearing (i.e., auditory thresholds below 30 dB HL across 250-8000 Hz), implying that the primary insult is centrally located and not at the level of the cochlea.¹⁵ One functional aspect of hearing that is likely to be disrupted in service members with TBI-S is temporal precision. Temporal precision allows for the timing of auditory events to be accurately encoded and is critical for skills such as spatial listening and binaural hearing. In particular, binaural hearing exploits very subtle timing differences (“interaural phase difference” [IPD]) between right and left ears to estimate the position of a sound source. Given that the neural substrate of binaural hearing decussates and crosses midline at the level of the brainstem, midbrain, and cortex through long axonal tracts, concussive blasts may be especially disruptive to this aspect of auditory processing.^16–18 A rapid, objective test of binaural hearing may thus be critical for assessing service members in training or the battlefield who have suffered repetitive concussive or blast TBI. With this in mind, AEPs at the level of the brainstem (envelope following response, frequency following response) and cortex (acoustic change complex [ACC]) using a binaural IPD paradigm in service members experiencing a TBI provide a critical tool for diagnosis of persistent TBI-S through the implementation of objective electrophysiological biomarkers. It is hypothesized that in persistent TBI-S, these biomarkers would not show recovery of temporal encoding, but rather reflect long-term damage to the central auditory pathways.

The visual system is similarly impacted by brain injury. Different structures demonstrate varying susceptibility to trauma and are therefore studied optimally using different tools. The central visual motor system (mesencephalon, paramedian pontine reticular formation, medial longitudinal fasciculus, etc.) is acutely affected by a mild traumatic brain injury (mTBI) and best studied using quantitative tools such as the video-oculogram (VOG). Oculomotor and vestibular deficits have been associated with both acute and chronic concussion injuries.^3,19–24 The vestibular–oculomotor system is currently assessed in the field using nonquantitative tests. Both the electronystagmogram and VOG are tools to objectively assess the status of an oculomotor system. Depending on the source, 20% to 40% of patients having experienced an mTBI show deficits in smooth pursuit, tracking, and pursuit initiation.²⁵

The evolution of an mTBI event to CTE is best studied using optical coherence tomography (OCT). Optical coherence tomography angiography is sensitive to perfusion abnormalities in the superficial plexus (retinal nerve fiber layer [RNFL]) of the retina and has been shown to be abnormal in several neuropathies including Alzheimer’s disease and Parkinson’s disease.²⁶ Furthermore, RNFL thinning has been reported in CTE.^27,28,29 We have developed a new imaging modality—scattering angle resolved (SAR)-OCT—that is also sensitive to retinal perfusion abnormalities and retinal ganglion cell degeneration.³⁰

As described above, each of these modalities provide potential in the differentiation between transient and persistent TBI-S. The authors anticipate that the aggregate of these modalities will provide a “signature” that will differentiate persistent TBI-S from transient episodes.

METHODS

The authors are advancing studies of AEP, OKN, and OCT measures and preparing to fuse the data using advanced computational methodologies employed at the University of Texas at Austin Advanced Research Laboratories in signals fusion. Just as signals from multiple naval or air sources might be fused to provide a more accurate operational picture, the authors seek to test how fusing multiple independent test data results can be more predictive of persistent TBI-S. The resulting array of data is expected to correlate to other more advanced tests such as magnetic resonance imaging or blood panels. The authors believe that further study and analysis will lead to validated quantitative indicators for continued training or deployment. It will likely take further work and analysis to generate validated field-ready indicators. Studies of recently concussed patients in hospital settings are proposed to develop data for further fusion analysis. Where possible, field studies of military personnel in training may also yield similar data.

AEPs

All AEPs will be recorded using a 64-channel high-density EEG Quik-Cap and a Neuroscan Synamps2 System (Compumedics, Inc.). Contact impedance will be maintained at ≤5 kΩ for all electrodes throughout the recording session. Click-evoked ABR will be recorded in response to rarefacting clicks (100 μs) presented at a rate of 31.1/s at 80 dB SPL. Responses are to be high-pass filtered at 100 Hz, epoched with a −5 to 12 ms window (re: stimulus onset), and averaged online. Right and left ears are to be tested independently. The results will be compared with a normative database, using 95% confidence intervals, to assess normalcy. Interaural phase difference brainstem and cortical responses will be elicited by an “in phase” condition, where stimuli will be presented in an identical fashion in both ears throughout the duration of the trial. Stimuli will also be presented in an “out of phase” condition, with pure tones beginning “in phase” for the first 400 ms of the stimulus. At the zero-amplitude instance of the amplitude modulation cycle occurring at 400 ms, the carrier tone IPD will change to 180° by inverting the tone polarity in the right ear (Fig. 2) in individuals with TBI. Stimuli will consist of 800 ms AM tones presented at a rate of 1/s with 100 ms pre- and poststimulus epochs. Presentation order is randomized by carrier frequency (500, 750, 1000, and 1250 Hz) and trial type (“in phase” or “out of phase”). Five hundred sweeps are to be acquired for each stimulus in a continuous EEG data file band-pass filtered at 0.1 to 3000 Hz. Both ABR and ACC responses to these stimuli will be assessed, providing a measure of binaural processing from the brainstem through the cortex.

FIGURE 2.

A cortically generated acoustic change complex (ACC) in response to a 180° interaural phase difference (IPD) change is seen on the right side of the second row (black arrow). In the top panel of the third row, envelope following response (EFR) waveforms to the 80 Hz AM envelope are shown. The EFR waveforms become desynchronized or less phase-locked (black arrow) when the IPD occurs; this phenomenon can be seen at 80 Hz in the time–frequency domain (white arrow) when the same data are plotted in terms of phase locking strength across frequency. Frequency following response (FFR) waveforms to the 500-Hz carrier frequency are plotted in the top panel of the fourth row. Like the EFR, FFRs become desynchronized following the IPD.

OKN

We have developed a low-cost but reliable VOG system based on the Android platform. The patient will view a drifting pattern of bars generated similar to the OKN drum app available through iTunes. A normal VOG and the Android phone platform are shown in Figs. 3a and 3b. The test will start by showing the user an OKN drum. Eye movement and pupil diameter will be recorded. The video will be processed to derive the horizontal displacement and diameter of the pupil over time.

FIGURE 3.

(a) Normal video-oculogram (VOG) recorded with the portable Android phone platform. X axis full scale = 7 seconds, Y axis full scale = 20° visual angle. The VOG system records eye movement by tracking the movement of the pupil using only the front-facing camera on an Android phone. It is portable and field-deployable. The device is also capable of recording the pupil diameter. The user runs the application and inputs the desired settings: spatial and temporal frequency. The device is inserted into a modified virtual reality (VR) headset. When the user is ready, a button is pressed on the VR headset and the test begins. (b) Photo of the VOG system developed in our laboratory with an example of an operator being examined.

The OKN test is an instrumentation to quantify oculomotor function. That test is useful to measure smooth pursuit, tracking, saccadic velocity, and pursuit initiation. The OKN test also measures nystagmus amplitude and direction. The OKN test is useful for measuring longitudinal changes in an individual with oculomotor dysfunction. For example, resolution of nystagmus would indicate recovery from an acute injury. There is no single parameter of the OKN measurement that is diagnostic of “persistent TBI”; however, abnormalities in longitudinal measurements that are present initially and fail to improve over time would indicate persistent oculomotor injury.

The video-oculogram measured with the portable Android cellphone has been validated against the electro-oculogram. They both measure rotation of the globe. The electro-oculogram is a clinical tool used by researchers and clinicians to assess oculomotor function and its usefulness has been well documented in the literature.

SAR-OCT

An optical approach to detect neurodegeneration of the RNFL in the human eye was developed.³⁰ The hypothesis of these studies was that early neurodegeneration could be detected by measuring the RNFL birefringence. A decrease in retinal light reflectivity during neurodegeneration was observed and the light reflectivity decrease was more easily detected than decreases in RNFL birefringence.³¹ A modified instrument, SAR-OCT was developed to take advantage of these observations.^30,32 To test the hypothesis that SAR-OCT can detect neurodegeneration, a murine SAR-OCT system was constructed and tested on a hypoxic murine model.³²

The SAR-OCT instrument is currently being tested in a triple murine knockout model of amyloid and tau pathology. Twenty mice with genetically induced Alzheimer’s have been imaged in these studies together with 12 control animals over a 1-year time period. The results indicate that SAR-OCT detects scattering differences between Alzheimer’s and control mice and that the differences were apparent at the first time point measured.³³

Retinal ischemia-reperfusion injury is measured by transient elevations of intraocular pressure that is known to induce neuronal damage. The quantitative changes in L/H2 are measures of neuronal injury. The Burr Type XII distribution is used here to parameterize the distribution of L/H2 values in any volumetric region of the retina. It was found that the distribution of L/H2 values in the retina fit nicely to a Burr Type XII distribution (following equation).³²  

Here α is a scale parameter (in the x-dimension), C and K are shape parameters, and x is the L/H2 ratio.

To detect changes in scattering angle, the distribution of L/H₂ ratios is traced over time by fitting the L/H₂ distributions to a Burr Type XII distribution and plotting temporal variation of the C parameter. Probability density functions are reported for L/H₂ for three groups of retinal layers: RNFL, ganglion cell layer, and outer retina. Image analysis includes RNFL thickness, total retinal thickness, volumetric blood flow (optical coherence tomography angiography), ganglion cell density, and reflectivity index.

Scattering angle resolved-OCT has not been studied in humans with TBI. Scattering angle resolved-OCT is a new clinical tool to assess scattering angle diversity in the RNFL. There are reports of decrease RNFL thickness in humans with TBI.³⁴ Since changes in scattering angle diversity occur before decrease in RNFL thickness, it is hypothesized that SAR-OCT might be an earlier biomarker for optic nerve damage than conventional OCT.

RESULTS

Preliminary AEP results in typical individuals indicate that brainstem (envelope following response [EFR]) and cortical (ACC) evoked potentials respectively show strong onset and periodic phase locking components in the “in phase” condition, indicating that timing differences between ears did not occur. Conversely, in the “out of phase” condition, stimulus timing between right and left ears is altered midway through the stimulus, resulting in changes in EFR phase locking at the level of the brainstem. Similarly, the “out of phase” condition elicits a second AEP peak or ACC at the level of the cortex. Both results indicate that phase changes have been successfully encoded from the brainstem through the cortex. It is therefore hypothesized that in the TBI-S population, these biomarkers will be significantly altered or absent, reflective of trauma to the central auditory pathways. The resulting data will be captured in trials where OKN and OCT data is also obtained to allow for fusion of the data to determine patterns of signatures correlating with other diagnostic indicators of TBI-S.

Optokinetic nystagmus has not been routinely measured in service members with mTBI. Protocols for data acquisition and analysis have been developed. All OKN data will be sent to a secure cloud-based storage for further study. The spatial and temporal frequency of the bar pattern can be varied to detect phase-locking to the drifting pattern under differing stimulus conditions. Patients with persistent TBI-S are expected to produce data reflecting changes in OKN.

The first step in the statistical analysis of the OCT contribution to fused data is to parameterize the RNFL, ganglion cell layer, and outer retina in four peripapillary quadrants and the macula region. Maps of the C parameter, GC cell density, reflectivity index, and volumetric vascular density are computed. A priori, which of the parameters (and at what retinal locations) best identifies the effects of CTE is unknown. Our hypothesis is that the retina will change after the brain injury. As subject data are acquired in sufficient numbers to allow for artificial intelligence to model baseline measures versus postrepetitive or blast exposures, we expect a model signature of predictability will emerge.

DISCUSSION

This manuscript reflects the hypothesis that the etiology of TBI-S that could lead to CTE may be the result of both repetitive overpressures and single blast exposure; and even repetitive concussive and sub-concussive events over time. Research of persistent TBI-S reveals how previous microbleeds at the depths of the sulci were hard to detect without autopsy. New methods and research reveal how audio and visual testing can measure TBI-S and suggest brain localities for further study. The authors’ hypotheses are supported by the following discussion while reviewing the results of research in our laboratories in concert with modeling.

Since current AEP and OKN devices are large and require clinical settings for reproducible results they have been almost exclusively used in hospital or laboratory settings. However, our university laboratories are in development of the auditory and visual testing tools containing further miniaturization, portability and potential field use. Using smartphone devices and potentially smaller footprint auditory measurement devices makes it possible to test more service members outside clinical settings for measurements and relay that data to collection for analysis and potential fusion for the development of the proposed fused baseline signature model early in the service member’s career.

Several AEP-based equipment manufacturers currently provide portable systems, but these are not at the stage necessary for successful field use. It is feasible and necessary to make these systems smaller and field-deployable to provide immediate testing for possible TBI-S symptomatology through AEP biomarkers. The needs include combining amplifier and recording hardware into a handheld device, as well as a user-friendly interface application. The development of portable military specification measuring devices that provide objective and quantitative data are obtainable with dedicated effort for field use by trained field medics. Moreover, if deployed simultaneously, we anticipate “signature” standards of the fused data can result in rapid diagnosis of persistent TBI when compared statistically to previously acquired baseline data.

All of the technologies and developments described here are ready to advance to further research. The basic principles of AEP, OKN, and OCT are well understood. Overall, if integrated simultaneously, fused data may provide significant, objective biomarkers resulting in rapid diagnosis of persistent TBI-S. The AEP requires about 40 minutes to acquire on a given individual. The OKN measures are accomplished in 5 minutes. We anticipate the development of prototypical eventual field hardened devices to move forward in 2020.

CONCLUSION

The authors surmise that the application of multiple tests through auditory and visual testing could advance our understanding of overpressures on the development of TBI-S and potentially CTE. The basic principles of AEP, OKN, and OCT are well understood. Large data sets of normalized data in the age range of 17 to 40 (range of our interest in our active service members) remain to be acquired but is easily obtainable. The diagnostic technologies (AEP, OKN, and OCT) currently exist at laboratory and clinical dedicated sites (not portable) and require refinement on discrimination in data analysis. What is required is fusion of the data from these neuroelectrophysiological technologies and continued development of portable and field-capable testing instrumentation deployable in training and potentially kinetic environments. The authors recommend further research to validate these hypotheses and correlate results through longitudinal studies.

FUNDING

The funding was provided by the University of Texas to S.S. (NIH K01DC017192-01) and H.G.R. (NIH T32EB007507) and the University of Texas System Neuroscience and Neurotechnology Research Institute.