Risk assessment models are developed to estimate the probability of brain injury during head impact using mechanical response variables such as head kinematics and brain tissue deformation. Existing injury risk functions have been developed using different datasets based on human volunteer and scaled animal injury responses to impact. However, many of these functions have not been independently evaluated with respect to laboratory-controlled human response data. In this study, the specificity of 14 existing brain injury risk functions was assessed by evaluating their ability to correctly predict non-injurious response using previously conducted sled tests with well-instrumented human research volunteers. Six degrees-of-freedom head kinematics data were obtained for 335 sled tests involving subjects in frontal, lateral, and oblique sled conditions up to 16 Gs peak sled acceleration. A review of the medical reports associated with each individual test indicated no clinical diagnosis of mild or moderate brain injury in any of the cases evaluated. Kinematic-based head and brain injury risk probabilities were calculated directly from the kinematic data, while strain-based risks were determined through finite element model simulation of the 335 tests. Several injury risk functions substantially over predict the likelihood of concussion and diffuse axonal injury; proposed maximum principal strain-based injury risk functions predicted nearly 80 concussions and 14 cases of severe diffuse axonal injury out of the 335 non-injurious cases. This work is an important first step in assessing the efficacy of existing brain risk functions and highlights the need for more predictive injury assessment models.

The majority of injuries were at the cranio-vertebral junction, indicating that the inertial head mass caused a tensile loading mechanism to the cervical spine. These data may be used in conjunction with finite element modeling to estimate risks to the human population. The most direct application in the automotive environment could be to the well-restrained child. The N neck injury criteria, currently based on data from piglet studies, could also benefit because the NHP is a more accurate human surrogate. These types of tests are likely to never be repeated and will form an upper bound of tolerance information valuable to safety system designers.

This work presents some of the challenges and solutions identified when using a historical collection of high-speed film from human research volunteer and anthropomorphic test device sled tests to define corridors for head flail kinematics. Challenges were related to film “jitter” and a lack of information about timing, test setups, and camera position. The challenges were addressed for a set of 939 tests using digitized high-speed film and minimal additional information for calculating head displacement, scaling, and translating to the seat origin. The method produced realistic head displacement data from tracked phototarget data. For the scaling method, the expanded uncertainty of measuring the phototarget size ranged from ±2.6 to ±3.7 pixels at a 95% confidence level. For the seat origin translation method, the average difference between calculated and actual seat origins was 6.6±5.0 cm. The methods implemented in this work can inform future work leveraging tracked kinematic data from this and other historical biodynamics datasets, especially large datasets that require batch processing.

Introduction Accelerative events commonly expose military pilots to potentially injurious + Gz (axial, caudal to cranial) accelerations. The Naval Biodynamics Laboratory exposed nonhuman primates (NHPs) to + Gz loading in two subject orientations (supine or upright) to assess the effect of orientation and accelerations associated with injury at accelerations unsafe for human participation. Materials and Methods Archived care records, run records, and necropsy and pathology reports were used to identify acceleration-related injuries. Injuries were categorized as central nervous system (CNS), musculoskeletal (MSK) system, or thoracic (THR). The occurrence of injuries relative to corresponding peak sled acceleration (PSA) and subject orientation were compared. A t-test was applied within each injury category to test for a significant difference in mean PSA between orientations. Results For all 63 + Gz runs conducted, PSA ranged between 6 and 86 G. Of these runs, 17 (6 supine, 11 upright) resulted in acceleration-related injury. The lowest PSAs associated with injury for supine and upright were 69.8 G and 39.6 G, respectively. Individual injury occurrences for supine runs (CNS [n = 5], MSK [n = 6], and THR [n = 6]) occurred at/above 75.7 G, 69.8 G, and 69.8 G, respectively. For upright runs, injury occurrences (CNS [n = 3], MSK injuries [n = 9], and THR injuries [n = 6]) occurred at/above 60.1 G, 39.6 G, and 50.5 G, respectively. The applied t-test showed significant difference between the mean orientation accelerations within each category. Injuries to supine NHPs included compression fracture, organ damage, brain hemorrhage, spinal cord hemorrhage, cervical hemorrhage, paresis/paraplegia, and THR bruising, whereas injuries to upright NHPs included compression fracture, organ damage, spinal cord hemorrhage, paresis/paraplegia, THR bruising, and difficulty breathing. Conclusions Axial loading to supine occupants produced more CNS injuries, whereas upright produced more MSK injuries. Both orientations produced equal THR injuries. NHP injuries reported reflected those reported following human + Gz acceleration events, highlighting the importance of orientation during acceleration to mitigate injury for next generation equipment design and testing.