It has been well established that ischaemia can result in secondary injury to the brain following trauma. While such ischaemia has been imaged, it has been difficult demonstrate its physiological significance. The aim of this research was to use diffusion weighted magnetic resonance imaging (DWI) to characterise the patterns of cytotoxic and vasogenic oedema early after head injury and correlate changes with regional physiology, imaged using 150 positron emission tomography (,sO-PET). Data from methodological developments carried out in the course of this research are presented These include the testing of MR compatibility of infusion pumps, optimisation of image processing routines, assessment of the validity of commonly used MR measures of tissue injury in the context of head injury, and an assessment of the test-retest reproducibility of DWI. Early DWI imaging in 30 patients with significant head injury (range 8 - 134 hours) revealed a characteristic contusional morphology, with a haemorrhagic core and concentric rings of vasogenic and cytotoxic oedema. In the regions studied, the integrated volume of pericontusional oedema was over three times the volume of the central core. An analysis across patients, although confounded by interindividual variation, suggested that this pericontusional oedema increased in size with time from injury. Correlation with electron microscopy suggested microvascular ischaemia as a mechanism for these changes The physiological correlates of the ADC changes described above were investigated in a subset of nine patients with ,sO-PET. The contusion core showed significant reductions in cerebral blood flow (CBF), oxygen extraction fraction (OEF) and cerebral oxygen metabolism (CMR02), while the region of vasogenic oedema only showed significantly reductions in CMR02. Other studies explored the use of dynamic DWI to assess the impact of hyperventilation on ADC changes around contusions. The implications of these findings are discussed and further research directions explored.

No description available.

Impact induced injury to the human head is a major cause of death and disability; this has driven considerable research in this field. Despite this, the methods by which the brain is damaged following non-penetrative (blunt) impact, where the skull remains intact, are not well understood. The mechanisms which give rise to brain trauma as a result of blunt head impact are frequently explored using indirect methods, such as finite element simulation. Finite element models are often created manually, but the complex anatomy of the head and its internal structures makes the manual creation of a model with a high level of geometric accuracy intractable. Generally, approximate models are created, thereby introducing large simplifications and user subjectivity. Previous work purports that blunt head impacts of short duration give rise to large dynamic transients of both positive and negative pressure in the brain. Here, three finite element models of the human head, of increasing biofidelity, were employed to investigate this phenomenon. A novel approach to generating finite element models of arbitrary complexity directly from three-dimensional image data was exploited in the development of these models, and eventually a highly realistic model of the whole head and neck was constructed and validated against a widely used experimental benchmark. The head models were subjected to a variety of simulated impacts, ranging from comparatively long duration to very short duration collisions. The dynamic intracranial pressure response, characterised by large transients of both positive and negative pressure in the brain, was observed following short duration impacts in all three of the models used in this study. The dynamic intracranial response was also recorded following short duration impacts of high energy, involving large impact forces, which were deemed to be realistic representations of actual impact scenarios. With the aid of an approximate analytical solution, analysis of the simulations revealed that the dynamic response is caused by localised skull deflection, which induces flexural waves in the skull. The implications of these magnified pressures are discussed, with particular regard to the potential for intracranial cavitation.