Field and analytic observations of impact brain injury in fatally injured pedestrians.

To develop more effective head protection against impact injury, maximum levels of mechanical impact or injury tolerance criteria, or both, should be specified for particular levels of injury and for particular structures of the brain. By using a development of an existing very simplified model of the head-vehicle impact for pedestrians we were able to make estimates of the peak angular acceleration and change in angular velocity of head impacts for fatally injured pedestrians. This model also enabled us to examine the relationship between the parameters of the impact, and the critical strain curves for brain injury proposed by Margulies and Thibault (1992). It was found that the offset of the impact from the center of mass of the head was a major influence, and, in addition, in impacts with a combined head/vehicle stiffness above 130 kN/m, the head impact velocity and change in head angular velocity were important, whereas for impacts with lower stiffness, the stiffness of the impact structure and hence, peak angular acceleration, were the major influences. Transformed into the frequency domain, the 130 kN/m region corresponds roughly to a harmonic of the natural frequency of the brain and skull, and the change in behavior may be related to decoupling of the skull and brain at impact. In 12 cases of lateral head impact, all but one case with visible injury in the corpus callosum were found to lie close to or above the 10% critical strain curve. Despite the very wide error limits around each data point, there is sufficient consistency between the field observations of brain injury and the analytic findings to suggest that the 10% critical strain curve represents a threshold for brain injury, expressed in terms of peak angular acceleration and change in angular velocity.

A method of estimating linear and angular accelerations in head impacts to pedestrians.

In order to investigate the relationship between impact to the head and brain injury, we have developed a method, using information obtained from reconstruction of the collisions, of estimating the peak linear and angular accelerations of the head for pedestrian impacts on a vehicle. This information includes the location of the impact on the head, the impact velocity of the head, and the stiffness of the struck surface. In developing the method we assumed that the velocity of the head on striking the vehicle was the same as the velocity of the vehicle itself, that the force vector was normal to the surface of the skull, that the force-deflection curve characterising the combined response of the impacted surfaces was linear, and that the kinetic energy of the head immediately prior to impact was converted into strain energy in deforming the head and the vehicle structure. Only the loading phase of the impact was considered, there was no assumption of an elastic unloading phase. Using cadavers, the validity of these assumptions and hence the usefulness of the method were tested by comparing the estimates of peak linear acceleration with the results of 18 pedestrian-vehicle impact reconstructions. On average, the method underestimated the experimental values by about 15%, with a range of +/- 35%. The results from the application of this method are currently being used to study the relationship between the magnitude and direction of the impact to the head and the distribution and severity of the brain injury resulting from actual collisions.

Brain injury patterns in fatally injured pedestrians.

To study the relationship between the severity of impact to the head and the severity and distribution of injury to the brain in fatally injured pedestrians, events in vehicle-pedestrian collisions were reconstructed to determine the peak linear and angular acceleration sustained by the pedestrians' heads. The nature and distribution of injuries to the brain were determined by neuropathologic examination of coronal sections of the brain. Study of 13 cases with occipital impacts and 18 with lateral impacts showed that the brain appeared to be more susceptible to injury from lateral impacts. The frontal and temporal regions appeared to be more susceptible to injury at low accelerations in occipital impacts, providing an explanation for "coup" and "contrecoup" injuries. For occipital impacts, a positive relationship was found between linear acceleration and the extent of injury to the brain, suggesting that there was a threshold for observable and concussive brain injury at about 1500 m/s2 peak linear acceleration. These findings are important for the development of measures for preventing brain injuries.