A finite element infant eye model to investigate retinal forces in shaken baby syndrome.

BACKGROUND: Shaken baby syndrome (SBS) is a form of abuse in which an infant, typically 6 months or less, is held and submitted to repeated acceleration-deceleration forces. One of the indicators of abuse is bilateral retinal hemorrhaging. A computational model of an infant eye, using the finite element method, is built in order to assess forces at the posterior retina for a shaking and an impact motions. METHOD: The eye model is based on histological studies, diagrams, and materials from previous literature. Motions are applied to the model to simulate a four-cycle shaking motion in 1 second with maximum extension/flexion of the neck. The retinal forces of the shaking motion, at the posterior eye, are compared to an impact pulse (60G) simulating a fall for a total duration of 100 ms. RESULTS: The shaking motion, for the first cycle, shows retinal force means at the posterior eye to be around 0.08 N sustained from the time range of 50 to 200 ms, into the shake, with a peak in excess of 0.2 N. The impulse, area under the curve, is 15 N-ms for 250 msec for the first cycle. The impact simulation reveals a mean retinal force around 0.025 N for a time range of 0 to 26 ms, with a peak force around 0.11 N. Moreover, the impulse for the impact simulation is 13 times lower than the shaking motion. CONCLUSION: The results suggest that shaking alone may be enough to cause retinal hemorrhaging, as there are more sustained and higher forces in the posterior retina, compared to an impact due to a fall. This is in part due to the optic nerve causing more localized stresses in a shaking motion than an impact.

Effects of initial seated position in low speed rear-end impacts: a comparison with the TNO rear impact dummy (TRID) model.

Injury-producing mechanisms associated with rear-end impact collision has remained a mystery not withstanding numerous investigations devoted to its scrutiny. Several criteria have been proposed to predict the injury-causing mechanism, but none have been universally accepted. The challenge lies in determining a set of testing procedures representative of real-world collisions, wherein the results obtained are not only the same as human testing, but remain consistent with various subjects and impact conditions. It is hypothesized that one of the most important considerations in the testing methodology is the effect of initial seated position (ISP) on occupant kinematics during a rear impact collision. This study involves two parts that evaluates the effects of ISP during rear-end impact. In the first part, head acceleration results of computer simulation using Hybrid III TNO rear impact dummy (TRID) are compared to physical impact testing (PIT) of humans. The second part focuses on the computer simulation using TRID to obtain different neck parameters such as NIC (Neck Injury Criterion), NIJ (Neck Injury Predictor), neck forces and moments to predict the level of neck injury such as whiplash associated disorder (WAD) during low speed rear-end impact. In PIT, a total of 17 rear-impact tests were conducted with a nominal 8-km/hour change in velocity to 5 subjects in four different seated positions comprising of a normal position (NP) and three out of positions (OOP). The first position was a NP, defined as torso against the seat back, looking straight ahead, hands on the steering wheel, and feet on the floor. The second position was a head flex position (HFP), defined as the normal position with head flexed forward approximately 20 degrees. The third position was a torso lean position (TLP), defined as the normal position with torso leaned forward approximately 10 degrees away from the seat back. Lastly, a torso lean head flex position (TLHFP), defined as the normal position with the head flexed forward approximately 20 degrees and torso leaned forward approximately 10 degrees. The head acceleration plots from PIT reveal that for the third and fourth positions (TLP and TLHFP) when the subject torso leaned forward, the peak head acceleration for the subject decreased and there was also a delay in reaching the peak. The Hybrid III-TRID anthropomorphic test dummy (ATD) was used in the same four different seated positions using computer simulation software MAthematical DYnamic MOdel (MADYMO 6.0) and the head acceleration results were compared to PIT. The comparison demonstrates that the Hybrid III-TRID ATD with MADYMO can be a reliable testing procedure during low-speed, rear-end impact for the four ISPs considered since the head acceleration plots deviated within the range of PIT head acceleration plots for different human subjects. This ensures that the second part of the study with neck injury using computer simulation results is a reliable testing procedure. It can be observed that MADYMO results have a greater error when compared to PIT when more than one OOP condition is employed as in TLHFP. All these observations would help in providing a tool to better understand the injury mechanisms and provide an accurate testing procedure for rear-end impact.