The purpose of this research was to conduct reconstructions of concussive and non-concussive impacts in ice hockey to determine the biomechanics and thresholds of concussive injury in ice hockey. Videos of concussive and non-concussive impacts in an elite professional ice hockey league in North America were reconstructed using physical and finite element model methods. Eighty concussive and 45 non-concussive events were studied. Logistic regressions indicate significant thresholds for concussion for linear/rotational acceleration and CSDM_10%. Impacts in ice hockey were mostly long duration events, longer than 15 ms. These results have significant implications for helmet standards and development to prevent concussion.

Collisions are a common cause of concussions. The nature of a collision impact is governed by the velocity of the striking players and the body parts involved. The purpose of this study was to describe the effects of velocity, mass, surface stiffness and location on peak linear and angular acceleration and strain. A variable-mass pendulum fitted with simplified impact surfaces representing the stiffness of an unprotected head, a protected (helmet) and a well-padded (i.e. shoulder pad) condition were used to impact a Hybrid III headform at four locations and velocities. Headform accelerations were input into a finite-element model to approximate peak strain in the cerebrum. Stiffness was most influential on peak linear and angular acceleration, and strain was most influenced by velocity. Interaction effects demonstrated that the 9 kg mass had consistently higher peak linear and angular acceleration and strain than the 15 kg mass between 3.0 and 4.5 m/s. Mass and velocity interactions may have important implications for age and size related concussion risk.

This study describes the interaction of impact parameters on peak head acceleration and strain variables for test conditions represented in sport. The Hybrid III 6-year-old child, 5th percentile female, and 50th percentile adult male headforms were subject to parametric tests using a monorail drop tower at four impact velocities (1.5, 3.0, 4.5, and 6.0 m/s), three surfaces (unprotected, protected/helmeted, and well-padded/mat), and four impact locations (frontal, sagittal, combined-plane motions, and a rotationally dominant motion). Scaled finite-element models of the brain were used to obtain peak strains. Regression analyses revealed that compliance produced the greatest increases in head acceleration, while impact velocity was for strain. Smaller headforms were associated with higher responses. Non-uniform trends for impact location were noted and are likely a result of localised headform properties interacting with velocity and compliance. These findings support the need for size-appropriate parameters in the design and development of head protection.

Concussion in American football is a prevalent concern. Research has been conducted examining frequencies, location, and thresholds for concussion from impacts. Little work has been done examining how impact location may affect risk of concussive injury. The purpose of this research was to examine how impact site on the helmet and type of impact, affects the risk of concussive injury as quantified using finite element modelling of the human head and brain. A linear impactor was used to impact a helmeted Hybrid III headform in several locations and using centric and non-centric impact vectors. The resulting dynamic response was used as input for the Wayne State Brain Injury Model to determine the risk of concussive injury by utilizing maximum principal strain as the predictive variable. The results demonstrated that impacts that occur primarily to the side of the head resulted in higher magnitudes of strain in the grey and white matter, as well as the brain stem. Finally, commonly worn American football helmets were used in this research and significant risk of injury was incurred for all impacts. These results suggest that improvements in American football helmets are warranted, in particular for impacts to the side of the helmet.

Multiple/repeated mild traumatic brain injury (mTBI) in young children can cause long-term gait impairments and affect the developmental course of motor control. Using our swine model for mTBI in young children, our aim was to (i) establish a reference range (RR) for each parameter to validate injury and track recovery, and (ii) evaluate changes in gait patterns following a single and multiple (5×) sagittal rapid non-impact head rotation (RNR). Gait patterns were studied in four groups of 4-week-old Yorkshire swine: healthy (n = 18), anesthesia-only sham (n = 8), single RNR injury (n = 12) and multiple RNR injury (n = 11). Results were evaluated pre-injury and at 1, 4, and 7 days post-injury. RR reliability was validated using additional healthy animals (n = 6). Repeated mTBI produced significant increases in gait time, cycle time, and stance time, as well as decreases in gait velocity and cadence, on Day One post-injury compared to pre-injury, and these remained significantly altered at Day Four and Day Seven post-injury. The gait metrics of the repeated TBI group also significantly fell outside the healthy RR on Day One, with some recovery by Day Four, while many remained altered at Day Seven. Only a bilateral decrease in hind stride length was observed at Day Four in our single RNR group compared to pre-injury. In sum, repeated and single sagittal TBI can significantly impair motor performance, and gait metrics can serve as reliable, objective, quantitative functional assessments in a juvenile porcine RNR TBI model.

Auditory and visually evoked potentials (EP) have the ability to monitor cognitive changes after concussion. In the literature, decreases in EP are commonly reported; however, a subset of studies shows increased cortical activity after injury. We studied auditory and visual EP in 4-week-old female Yorkshire piglets (N = 35) divided into anesthetized sham, and animals subject to single (sRNR) and repeated (rRNR) rapid non-impact head rotations (RNR) in the sagittal direction. Two-tone auditory oddball tasks and a simple white-light visual stimulus were evaluated in piglets pre-injury, and at days 1, 4- and 7 post injury using a 32-electrode net. Traditional EP indices (N1, P2 amplitudes and latencies) were extracted, and a piglet model was used to source-localize the data to estimate brain regions related to auditory and visual processing. In comparison to each group’s pre-injury baselines, auditory Eps and brain activity (but not visual activity) were decreased in sham. In contrast, sRNR had increases in N1 and P2 amplitudes from both stimuli. The rRNR group had decreased visual N1 amplitudes but faster visual P2 latencies. Auditory and visual EPs have different change trajectories after sRNR and rRNR, suggesting that injury biomechanics are an important factor to delineate neurofunctional deficits after concussion.