Modelling Infant Head Kinematics in Abusive Head Trauma

Abstract

Abusive head trauma (AHT) is a potentially fatal result of child abuse, but the mechanisms by which injury occur are often unclear. In this thesis, a novel computational framework for investigating head kinematics during AHT was developed using OpenSim (Delp et al., 2007), and was validated with kinematic measurements during shaking of an experimental phantom. The framework was used to investigate the biomechanics of AHT using model-based interpretation of animal shaking experiments, and computational studies simulating the shaking of human infants. The lamb was used as an in vivo experimental analogue of AHT. An OpenSim computational model of the lamb was developed and used to interpret biomechanical data from shaking experiments. Sagittal plane acceleration components of the animal’s head during shaking were used to provide in vivo validation of the computational framework. Results demonstrated that peak accelerations occurred when the head impacted the torso and produced acceleration magnitudes exceeding 200 m∙s-2. The computational model demonstrated good agreement with the experimental measurements and was able to reproduce the extreme accelerations that occur during impact. The biomechanical results demonstrate the utility of using a coupled rigid-body modelling framework to describe infant head kinematics in AHT. To investigate AHT in human infants, a novel probabilistic analysis of head kinematics during shaking was performed. A deterministic OpenSim model, incorporating an infant’s mechanical properties, was subjected to a variety of shaking motions. Monte Carlo analyses were used to simulate the range of infant kinematics produced as a result of varying both the mechanical properties and the type of shaking motions. By excluding physically unrealistic shaking motions, worst-case shaking scenarios were simulated and compared to existing injury criteria for a newborn, 4.5 months, and a 12 months infant. None of these cases produced head kinematics that exceeded previously estimated subdural haemorrhage injury thresholds. The results of this study provide no biomechanical evidence to demonstrate how shaking alone can cause the injuries observed in AHT, suggesting either that additional factors, such as impact, are required, or that the current estimates of injury thresholds should be interpreted with caution.