Head kinematics during shaking associated with abusive head trauma

Abstract

Abusive head trauma (AHT) is a potentially fatal result of child abuse but the mechanisms of injury are controversial. To address the hypothesis that shaking alone is sufficient to elicit the injuries observed, effective computational and experimental models are necessary. This paper investigates the use of a coupled rigid-body computational modelling framework to reproduce in vivo shaking kinematics in AHT. A sagittal plane OpenSim computational model of a lamb was developed and used to interpret biomechanical data from in vivo shaking experiments. The acceleration of the head during shaking was used to provide in vivo validation of the associated computational model. Results of this study 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 shown to be able to reproduce the high accelerations that occur during impact. The biomechanical results obtained with the computational model demonstrate the utility of using a coupled rigid-body modelling framework to describe infant head kinematics in AHT.

Introduction

Abusive head trauma (AHT) is a leading cause of traumatic brain injury in children (Duhaime et al., 1998, Parks et al., 2012a, Parks et al., 2012b) that is characterised by a triad of symptoms comprising of subdural and retinal haemorrhages and acute brain dysfunction (Harding et al., 2003). Shaking of infants by caregivers is believed to be a common feature of the abusive injury, but the mechanism of injury is unclear and there is debate about whether shaking alone is sufficient to elicit the injuries observed (Geddes et al., 2001, Geddes et al., 2003, Geddes et al., 2004, Punt et al., 2004, Squier, 2008).

To further the understanding of injury mechanisms, a number of animal experimental and computational models have been used. Early experiments were performed on primates to investigate injury thresholds during whiplash (Ommaya and Hirsch, 1971, Ommaya et al., 1968) and the injuries caused by translational and rotational head motion (Gennarelli et al., 1982, Ommaya and Gennarelli, 1974). These experiments were designed to investigate traumatic head injuries resulting from motor vehicle accidents in adults. Simple brain mass scaling was used to extrapolate insights from these adult primate data to children (Duhaime et al., 1987), but experimental work in piglets has shown that age-dependant material properties of brain tissue may have a large effect on biomechanical injury thresholds (Thibault and Margulies, 1998). A suitable animal model that exhibits all features of abusive head trauma in infants has not yet been described. Recently, repeated manual shaking experiments in lambs were shown to produce microscopic damage (Finnie et al., 2010, Finnie et al., 2012) but macroscopic bleeding, such as is reported in infants clinically, has not been observed.

Coupled rigid-body computational models of the paediatric head and neck have been developed to investigate car crash injuries (Dibb et al., 2014) and have been used to describe head kinematics during infant shaking (Wolfson et al., 2005). However, no in vivo shaking validation was performed by Wolfson et al. who instead relied upon mechanical surrogates (Cory and Jones, 2003, Duhaime et al., 1987). The unification of computational and experimental models is necessary to justify the use of a rigid-body modelling framework to describe the head kinematics of AHT and to assist in providing a mechanistic link between the input shaking motion and head injury.

The primary objective of this study was to develop a coupled rigid-body computational model of a lamb that is based upon in vivo head kinematics measurements. In reproducing the kinematics of lambs' heads during manual shaking, the results address the hypothesis that the modelling framework is an appropriate tool for investigating the kinematics of AHT in human infants.