Abstract
This thesis is the result of a PhD study at the University of Surrey in collaboration with The Institute of Cancer Research and Royal Marsden Hospital NHS Trust. The objective was to investigate and develop quantitative diffusion-weighted MR imaging techniques suitable for abdominal imaging. Diffusion-weighted imaging is not a new technique. However, until recently it has been mainly restricted to the head and extra-cranial diffusion imaging is still considered an experimental technique. A possible candidate for extra-cranial imaging is the Burst-diffusion technique as it provides geometrically correct images free from susceptibility and chemical shift artifacts. In a first clinical study promising results were obtained. We showed a clear correlation between the MR based apparent diffusion coefficient measured in rectal adenocarcinomas before cancer treatment and treatment outcome. In a more thorough analysis of the data based on the IVIM model, it was shown that the tumour response is in fact linked to the diffusion fraction of the tumour. The drawbacks of the original sequence used, namely low SNR, uni-directional diffusion-weighting and diffusion-weighting that depends on the chosen field-of-view, are discussed in detail and possible solutions proposed. This led to the development of a new type of Burst-diffusion sequence, which was tested with a newly developed diffusion-phantom based on aqueous and DMSO-based MnCl2-solutions in varying concentrations. The major problem of the new Burst-diffusion sequences, however, was their stronger sensitivity to motion. Currently, the fastest MR method is echo-planar imaging, with acquisition times in the order of 100 ms per b-value image. In EPI, well-known problems make a single-shot approach not feasible for extra-cranial imaging. Therefore, multi-shot EPI was investigated for diffusion-weighted imaging. These experiments led to the development of 2-shot interleaved steady-state GE- and SE-EPI sequences. The technique, however, will require further improvement, which was not possible with the hardware given. The last set of experiments compared the Burst and EPI sequences developed during this PhD with the techniques HASTE and PSIF. Experiments were performed on a phantom and in vivo on the brain and the pelvis. Each sequence has its pros and cons over the other techniques, with the HASTE sequence being the best option for the hardware used, assuming single-slice acquisition, and the particular PSIF sequence used being the least feasible one.