On the Shaken Baby Syndrome : Biomechanical Experiments and Numerical Simulations

Abstract

Shaken Baby Syndrome (SBS) is a brutal result of child abuse with severe and often fatal outcome. This Master's thesis describes some results from the analysis of acceleration data in a biomechanical experiment of dummy shaking.

An anthropometric baby crash test dummy (Q0 from FTSS) was obtained. The head accelerations at six points in the head and at the top of the head (the vertex) were registered as test volunteers shook the dummy freely.

Realistic modeling of the infant neck biomechanics is complex. Previous biomechanical studies have used dummies with necks possessing simple one-way hinge joints with no resistance, or short rubber necks. The measured peak accelerations have been low-scale, and researchers have questioned whether violent shaking indeed may be the true mechanism of injury in SBS. The Q0 dummy is supplied with an elastic neck enabling motion in all directions, and four metal plates simulating some of the cervical vertebra.

We have studied the peak acceleration shaking session, analyzed the acceleration data, and compared the results with literature. We have also developed a simple linear elastic Finite Element (FE) baby head model, and computed the shear strain and displacement for the peak acceleration field. We have carefully studied the elastic properties of infant brain tissue, and suggest that the brain should be modeled compressible in shaking, due to longer time duration of acceleration than in impacts.

Previous biomechanical studies have not measured or computed 3D acceleration. There are several difficulties with 3D acceleration computations for small sized heads. In this thesis we have developed a new solution for 3D acceleration computation, studied the rigid motion formula written as a matrix exponential, and validated the vertex measurements in new ways.

In our model we measured a higher peak acceleration than reported in previous studies. We found the rotation axis to move continuously during shaking, a finding that needs to be further analyzed with regards to the localization of brain injuries sustained in victims of SBS. Preliminary analytical and numerical analysis of the vertex measurements showed the acceleration measurements to be presumably valid.

The FE model behaves reasonable compared with human post-mortem experimental data. The brain/skull displacement was 2.5 mm at our shaking peak acceleration field, and this is too low due to the limitations in our model. The FE model should be further developed, and especially the fluid layer (CSF) surrounding the brain needs a better simulation.

The thesis describes the magnitude, direction and additional aspects of a head acceleration data set obtained from an experimental dummy shaking experiment. The analyses may shed light on the mechanism of injury in victims of SBS and provide a base for further research.