Couper, Zachariah Sol (2007). THE POTENTIAL FOR PEDIATRIC TRAUMATIC BRAIN INJURY DUE TO SHAKING PhD Thesis, School of Engineering, University of Queensland.

Attached Files (Some files may be inaccessible until you login with your UQ eSpace credentials)
Name Description MIMEType Size Downloads
n01front_couper.pdf n01front_couper.pdf application/pdf 18.51MB 2
n02content_couper.pdf n02content_couper.pdf application/pdf 18.50MB 2
Author Couper, Zachariah Sol
School, Centre or Institute School of Engineering
Institution University of Queensland
Publication date 2007
Thesis type PhD Thesis
Supervisor Dr Faris Albermani
Abstract/Summary Shaken Baby Syndrome (SBS) is understood to refer to a recurring and specific pattern amongst infants presenting with brain injury, characterized by retinal and extracerebral haemorrhage, and a lack of external trauma to the head and neck. While numerous cases of SBS have been diagnosed, its validity is a contentious issue on both biomechanical and medical fronts, primarily due to a lack of understanding of the loading-injury relationship of infant shaking, and the parameters which are deterministic to its nature. In order to address this lack, a series of finite element representations of a three month infant head were developed to utilise the results of physical testing with an anthropomorphic infant surrogate. Physical testing involved subjecting the infant surrogate to manually applied shaking confined to the sagital plane, in the anteroposterior direction. The 3D kinematics of the torso and head of the surrogate were recorded using a motion tracking system. The output history was converted to an ‘average shake’ profile, which in turn yielded translational and rotational sagital plane accelerations in the anatomic reference system of the head. This data is suitable for application to numerical models as boundary conditions, and constitutes a fundamental data set for future investigations. Brain matter material properties are a key component of finite element models of the head. Given the disparate body of research in this field, a detailed synthesis was conducted to determine appropriate representations for use in infant head models. This led to the proposal of a new infinitesimal strain viscoelastic model which was in good agreement with much of the past testing, and covered a much larger frequency range than has generally been used. Additionally, recently developed large strain models were proposed for use in future numerical modeling, given their superior ability to represent both shear and uniaxial test data for isotropic and transversely isotropic brain matter. Building upon the previous chapters, a series of 2D finite element models of the infant head were created in order to parametrically investigate the effects of cerebrospinal fluid (CSF) modelling technique, CSF thickness, brain matter stiffness, dissipation and nonlinearity, and material differentiation. A novel method based on Reynolds lubrication theory was used to represent the subarachnoid CSF. The results attest to the need to represent CSF through fluid rather than solid mechanics, particularly in the loading regime associated with shaking, as opposed to impact. Furthermore, it is evident that the brain-CSF interaction is highly dependant on both the subarachnoid CSF volume, and the thickness variations associated with protusions of the gyri. These protusions alleviate deep brain stress concentrations and hence aid injury mitigation. Differentiating between types of brain matter led to noticeable changes in the evolution of the brain-CSF interaction, outside of those caused by average stiffness changes. Inclusion of the pia mater was important from a local stress reduction perspective. These results aided the development of a 3D finite element model of the infant brain. A method allowing for the creation of more optimal meshes of the head with the capacity for case specific adaption was developed and applied to create the model. CSF was represented through static pressure equilibration in combination with a locally based squeezing resistance. The results of the simulation indicate that anteroposterior shaking will lead to specific patterns of brain matter motion, with a rotational squashing caused at both extremes of the shaking cycle, combined with translational acceleration effects which generate a tendency for the brain to cleave to the underside of the superior portion of the cranium. Significant contact between the brain matter and the cranium/membranes is caused at each squashing extreme, while high strains are also developed in the corpus callosum and brainstem/cerebrum/cerebellum connections. Taking into account the repetition of these strain levels, it appears likely that focal axonal injury would be generated at these locations. The relative motion between the cranium and brain matter is substantial, and sufficient to overstretch many of the bridging veins, particularly those located superiorly, indicating a capacity for the development of subdural hematomas. Taken together, these results indicate the act of anteroposterior shaking of an infant is sufficiently traumatic to cause injuries to the contents of the cranial vault, in the absence of blunt trauma. The results of this thesis are suitable for consideration by medical practitioners. The partial predication of Shaken Baby Syndrome diagnosis on subdural hematoma development appears to be justified; in addition, improvements to the confidence of SBS diagnosis could be aided by the targeted detection of focal axonal injury. However, further work is required to improve the fidelity and performance of the numerical models, with the eventual aim of establishing a detailed correspondence between an applied head insult and the resulting injury pattern.

Citation counts: Google Scholar Search Google Scholar
Created: Fri, 21 Nov 2008, 15:37:49 EST