Head Injuries in Multiple Domains

Abstract

Head and brain injuries are significant public health issues that affect individuals worldwide and encompass a wide range of injury types and severities. Traumatic head and brain injuries are common injuries observed in sports and motor vehicle crashes (MVC). Specifically, within MVCs they are a common cause of death and disability. Despite decades of dedicated research, the fundamental underlying causes of traumatic brain injuries remain poorly understood. Studying head injuries in vivo requires reliable instrumentation systems which can accurately measure head kinematics under potentially injurious conditions. Head kinematics obtained from reliable instrumentation can be used to study brain response in silico and can support the development of improved injury criteria and injury risk functions. Additionally, with enhanced instrumentation our ability to investigate differences in TBI across different populations (e.g. biological sex or age) would also be greatly enhanced.Head injuries were studied across multiple domains in this study. First, relative risk of fatality and injury for female vehicle occupants were examined. Data from multiple automotive safety and public health datasets (National Highway Traffic Safety Administration (NHTSA) Fatality Analysis Reporting System (FARS); Centers for Disease Control (CDC) Multiple Cause of Death (MCOD)) were linked to examine head injuries and overall differences in fatality rates in motor vehicle crashes. Matched cases were examined to determine the relative risk (R) of fatality or injury using a double pair comparison method. Young females (20s-40s) are at approximately 20% higher risk of dying in car crashes compared with males of the same age in matched scenarios. For example, 25-year-old female occupants in passenger car crashes were at 20% higher risk of fatality (R = 1.201; 95% CI 1.160–1.250) compared to 25-year-old males. This trend persisted across vehicle type, airbag deployment, seatbelt use, and number of vehicles involved in a crash. Similar trends were noted for matched crashes where head injury was the cause of death: 20-25 year old female occupants were 30% more likely to suffer fatal head injuries than males of the same age (R = 1.29; 95% CI 1.17-1.42). This increased risk was pronounced for young females in both fatal crashes and non-fatal crashes with severe injuries. Second, the biofidelity of in vivo wearable instrumentation systems used in the study of head kinematics was tested. Instrumented mouthguard systems (iMGs) are commonly used to study rigid body head kinematics across a variety of athletic environments. While iMGs rigidly fixed to anthropomorphic test device (ATD) headforms have demonstrated good correlation with reference systems in previous work, only one validation study focused on iMG performance in human cadaver heads. To assess iMG biofidelity, three unembalmed human cadaver heads were fitted with two instrumented boil-and-bite mouthguards (Prevent Biometrics and Diversified Technical Systems (DTS)) per manufacturer instructions, and were fitted with a properly-sized Riddell SpeedFlex American football helmet. Reference sensors were rigidly fixed to each specimen. Specimens were impacted with a rigid impactor at three velocities and locations. The Prevent iMG performed comparably to the reference system up to approximately 60g in linear acceleration, but overall had poor correlation (CCC = 0.39). The DTS iMG consistently overestimated the reference across all measures, with linear acceleration error ranging from 10-66%, and angular acceleration errors greater than 300%. Overall, neither iMG demonstrated consistent agreement with the reference system. While iMG validation efforts have utilized ATD testing, this study highlighted the need for cadaver testing and proper validation of devices intended for use in vivo, particularly when considering realistic (non-idealized) sensor-skull coupling, when accounting for interactions with the mandible and when subject-specific anatomy may affect device performance.Third, a population of Muay Thai martial arts athletes were instrumented with an in-ear wearable sensor system (DASHR) to better understand the epidemiology of head impact exposure during active sparring training. Martial arts athletes sustain many head impacts in a short period of time, and many martial arts forms focus competition victories on knock outs – the loss of consciousness or observable deficits in cognition and motor function due to repetitive impacts to the head. In this study, Muay Thai athletes were found to sustain 2.3 ± 1.8 head impacts per minute while sparring. Impact magnitudes were significantly higher during sessions when athletes wore headgear than during sessions that they did not, with linear accelerations 5.6g higher with headgear (Hodges-Lehman estimate of difference between medians; Mann-Whitney U-test p < 0.0001), angular velocities 1.6 rad/sec higher (p < 0.0001), and angular accelerations 710 rad/s^2 higher (p < 0.0001). While the use of headgear in martial arts has been shown to reduce accelerations during ATD drop testing, and to reduce the incidence of facial lacerations and contusions in competition settings, there is no consensus on the degree to which they protect against concussions and brain injuries. This study showed that during low magnitude impacts in a training setting, head kinematics are higher when headgear is worn than when it is not, potentially suggesting behavioral differences which offset any impact attenuation benefits the headgear offers. A large percentage of the impacts observed in this study (28%) fell below established recording thresholds used by most field-deployable wearables, suggesting a need for triggerless, continuous field recording systems. Finally, the efficacy of existing injury metrics and risk functions were tested assessed using 731 non-injurious impacts recorded in the Muay Thai population. Head impacts were assessed with 14 injury metrics and 30 injury risk functions. Injury risks were summed across all impacts for each metric to determine the number of expected injury events. Risk functions developed from reconstructed impact events overpredicted injury rates more than risk functions developed from field data. For example, Head Injury Criterion risk functions developed from reconstructed American football events predicted 45-126 concussions, compared to 0 concussions predicted by risk functions developed from collegiate football field data. While cumulative impact exposure may be an important factor in individual risk tolerances, only one measure (RWE) was related to injury outcomes, and predicted concussions for seven of nine athletes in this study where no concussions were observed. Overall, these studies highlight areas where injury biomechanics research has attempted to move forward without a reliable foundation to build upon. A lack of equitable cadaver testing has led to a gap in the biomechanical understanding of potentially fundamental underlying differences between males and females, reflected in sex differences in injury and fatality rates in motor vehicle crashes. Wearable head instrumentation systems which appear to perform well in controlled laboratory conditions are deployed without validation against biofidelic boundary conditions, significantly limiting their trustworthiness and usability. A large proportion of recorded head acceleration events using an in-ear sensing system fell below established trigger thresholds, suggesting a need for expanded data collection and continuous recording. The lack of cumulative exposure measures correlating to injury outcomes may, in part, be due to a lack of information in this low-level regime. Finally, established injury risk functions incorrectly classified many non-injurious head acceleration events, suggesting a need for new approaches to injury risk assessment using accurate in vivo kinematics as their basis.