The work presented in this thesis addresses some of the unresolved issues surrounding the use of PET for identifying and quantifying inflammation in atherosclerotic plaques. Six main questions were posed at the beginning the thesis. Firstly, does plaque FDG truly correlate with the degree of inflammation? Secondly can <i>in-vivo</i> FDG-PET imaging distinguish between inflamed and non-inflamed lesions in a small animal model of atherosclerosis? Thirdly, can the combination of FDG-PET and magnetic resonance (MR) imaging help in the assessment of “culprit” lesions in the carotid circulation in patients who have suffered a stroke? Fourthly, does the limited spatial resolution of PET lead to significant error (a phenomenon called partial volume error) when quantifying plaque FDG uptake in human carotid arteries? Fifthly, how reproducible is FDG-PET when it comes to quantifying FDG uptake in plaques. And finally, does PK11195, a novel synthetic receptor ligand, bind to macrophages and if so can it be used as an alternative to FDG for identifying highly inflamed plaques. The results presented in this thesis confirm a strong correlation between FDG uptake and the degree of macrophage infiltration in atherosclerotic plaques, thus supporting the use of FDG-PET as a technique for quantifying plaque inflammation. <i>In-vivo</i> quantification of plaque FDG uptake was found to be feasible in humans, is highly reproducible, and in combination with MR imaging could potentially be used to help assess and manage patients with atherosclerosis affecting large vessels such as the carotid artery. However, the partial volume effect can lead to significant quantification error, and needs to be taken into account when using this technique. Finally, although encouraging <i>ex-vivo</i> results were seen, there is currently no evidence to suggest that radiolabelled PK11195 is taken up into atherosclerotic plaques <i>in-vivo</i>. However further investigation is needed before ruling out the use of PK11195 for <i>in-vivo</i> imaging of plaque inflammation.

In this thesis, various alternatives to standard sensor systems are developed, with particular emphasis on improving the practicality of freehand 3D ultrasound in a clinical environment. The novel algorithms presented here are able to position a sequence of image frames qualitatively correctly even when the direction of out-of-plane motion is allowed to vary, which was not previously possible. Existing work on the subject of six-degree-of-freedom sensorless reconstruction found that it is possible to take unbiased measurements from a sensorless reconstruction, provided it is constrained to have only elevational motion and has a large proportion of fully developed speckle data. Alternatively, freehand motion and data from real tissue can be allowed at the expense of some large-scale drift in the frame positions, biasing the measurements. The approach of this thesis allows fully freehand motion and scans on any type of tissue, but consequently is found to have a drift error in the length and tilt of the sequence which can result in length measurement errors of 20% or more. A hybrid approach reintroduces position sensors, in combination with the image-based techniques. This thesis considers two sensors for use in the hybrid system. One of these is the Xsens MT9-B, which uses Micro-Electro-Mechanical Systems (MEMS) magnetometers, accelerometers and rate gyroscopes to measure orientations. This was found to be particularly unobtrusive having almost no effect on the scanning setup and protocol. A hybrid reconstruction using the MT9-B has an estimated overall length error of 5% of the sequence length and an orientation error of no more than 1°. A hybrid system using the MT9-B represents a significant step towards the goal of accurate anatomical measurements with an easy-to-use scanning system.

The purpose of the work described in this thesis was to investigate the potential of MRI for the longitudinal study of pharmacological activity within the central nervous system. For these investigations a robust physiological rat preparation, together with optimised imaging hardware, was developed for preclinical functional MRI studies at clinical field strengths of 2 Tesla. Graded asymmetric spin-echo echo planar imaging (ASE-EPI) was used to investigate changes in cerebral blood oxygenation induced by transient hypoxia in the anaesthetised rat. Increments of spin-echo asymmetry produced a linear increase in the sensitivity of data acquisition to the blood oxygenation level changes and revealed a heterogeneous response across four brain regions (the outer and inner neocortex, hippocampus and medial geniculate nucleus). In addition, the graded ASE-EPI protocol provided data pertaining to the vascular components that contributed to the observed signal. Having established optimal ASE-EPI imaging methods, the protocols were adapted to determine if pharmacologically induced metabolic changes were detectable in the anaesthetised rodent. To this end, the effects of two well characterised psychotropic compounds (MK-801 and mCPP) were studied. Time resolved statistical maps of significant signal change (<I>p</I> < 0.001) were produced for each brain slice using a rolling <I>t</I>-test to identify regions of interest for further analysis. The functional MRI signal closely mirrored the spatial pattern of both localised cerebral blood flow and glucose uptake, previously established by empirical autoradiographic methods. This work demonstrates that multiple time points of pharmacological activity can be mapped in a single subject using MR techniques.

No description available.

No description available.

Anti-vascular cancer therapy is based on experimental evidence that the growth of solid tumours relies on the development and continual expansion of a host-derived vascular network to feed the proliferating mass of cancer cells. Combretastatin A4 (CA4) has emerged as a promising anti-vascular drug as it specifically damages tumour vasculature, leading to extensive secondary cancer cell death, without significantly impairing healthy tissue perfusion. The aim of this research was to further our understanding of CA4’s mode of action, and in particular address the hypothesis that CA4 is active upon proliferating (angiogenic) but not quiescent blood vessels <i>in vivo.</i> Two experimental approaches using human tumour models xenografted in mice were adopted for this purpose. In the first, tumour vascular proliferation was limited within a CA4-senstivie tumour model using the anti-angiogenic drug, vascular endothelial growth factor (VEGF) receptor tyrosine-kinase inhibitor\SU. This was shown to significantly reduce the tumour’s vascular response to CA4 treatment as assessed by magnetic resonance imaging and spectroscopy. Secondly, a relatively CA4-resistant tumour model was genetically modified to overexpress VEGF, a major mediator of the angiogenic process. The vasculature of these tumours was shown to be sensitised to the vascular-damaging effects of CA4. The results of this research suggest that CA4 specifically targets angiogenic tumour vessels <i>in vivo.</i> This work provides a biological explanation for the drug’s specificity for tumour versus healthy vessels, and more specifically, can explain the wide variation in tumour response to this anti-vascular drug.

No description available.

No description available.

Apparent diffusion coefficient (ADC) measurements can be mapped over the feline cerebral cortex using diffusion-weighted echo planar imaging (DW-EPI). These were used to visualise and quantify the propagation of DS events and characterise their associated ADC changes. The present study established for the first time that the characteristics of SD are profoundly affected by the nature of its initiating stimulus. Thus, a sustained stimulus produced permanent MRI intensity changes local to the site of stimulus application. It initiated a single primary followed by multiple secondary SD events. All their ADC deflection amplitudes were similar and conserved throughout the course of their propagation across the cortex. The secondary events were separated by similar time intervals but propagated with lower velocities than the primary SD events. In contrast, a transient stimulus did not produce any permanent MRI changes in parallel with expectations following migraine aura. It initiated fewer events although SD activity nevertheless persisted well after stimulus withdrawal consistent with expectations from a regenerative process. Furthermore, ADC deflections produce by the primary SD events progressively declined with their propagation. The secondary SD events produced significantly smaller ADC deflections separated by progressively larger time intervals, but had propagation velocities similar to their preceding primary SD events. The effect on SD events of the 5-hydroxytryptamine [5-HT_1B/1D] agonist sumatriptan an agent established for the management of migraine and the novel benzoylamino benzopyran compound tonabersat were then compared. MRI made it possible to examine separately the actions of these compounds on the spatial extent and temporal characteristics of SD. Both compounds reduced the area affected by the primary SD event and the total period of SD activity. However, sumatriptan increased the number of SD events crossing the suprasylvian sulcus. In contrast, tonabersat reduced the frequency of SD events and in one case blocked all SD activity.

Dynamic Nuclear Polarization (DNP) is an emerging technique for increasing the sensitivity of Magnetic Resonance Imaging (MRI) in the liquid state. It has recently been applied to <i>in vivo </i>imaging of carbon metabolism: the spatial distribution of an injected hyperpolarized ¹³C-labelled molecule can be imaged in an intact living system, as well as the metabolites formed from it. This work demonstrates how this technique could have potential applications in medicine. ¹³C-labelled bicarbonate was hyperpolarized and the production of hyperpolarized carbon dioxide has been used to image tumour pH <i>in vivo</i> as well as the pH of the murine brain. An adaptation of this experiment allowed the spatial distribution of the enzyme carbonic anhydrase to be imaged <i>in vitro</i> and <i>in vivo</i>. Fumarate, a citric acid cycle intermediate, was hyperpolarized and its subsequent conversion into malate is shown here to increase with cellular necrosis <i>in vitro</i>; this was used as an early marker of response to chemotherapy both <i>in vitro</i> and <i>in vivo</i>. The metabolism of two other molecules is demonstrated: the <i>in vitro</i> metabolism of hyperpolarized ¹³C-labelled glutamine to glutamate as well as the metabolism of hyperpolarized ¹³C-labelled glutamate to α-ketoglutarate <i>in vivo</i>. Both of these present new ways to probe the citric acid cycle and the metabolism of glutamine may also act as a marker of cell proliferation. All of the molecules described here are endogenous and some are already administered intravenously into humans. There is therefore a realistic prospect that this technology can be translated into human imaging in the near future.