Diffusion MRI is a technique that is capable of providing unique contrast that is sensitive to molecular displacement motion at cellular and sub-cellular length scales. By sensitizing MR signal to the random motion of water molecule protons at a microscopic level (of the order of 5-20um), it is able to probe tissue microstructures in the brain such as axons, dendrites, glial cells, and extra-cellular spaces, in a manner that may provide valuable insights into tumor physiology.
Diffusion imaging is routinely acquired as part of the MR protocol for patients with brain tumors. However, the implications of the parameters being used are often not appreciated by the oncology community. This is especially true when applied to patients with high-grade glioma, where the lesion is highly heterogeneous and changes in diffusion parameters are due to a combination of treatment effects, edema and tumor infiltration. Although advanced diffusion models that aim to distinguish between different types of tissue have the potential for providing information that is complementary to conventional MR imaging, their application has been very limited due to their relatively long acquisition time.
These challenges have become the motivation for this thesis. We first explored the value of standard diffusion imaging methods in characterizing tumor response to therapy. By applying different ways of evaluating changes in the apparent diffusion coefficient (ADC) and examining their association with patient outcomes in clinical trials, we hoped to gain a better understanding of the physiological process behind the patterns of changes that occur, and improve the interpretation of the data obtained. The next step was to bridge the gap between advanced diffusion models and their clinical applications by using fast diffusion imaging techniques. This was achieved by optimizing the protocol for acquiring multiband diffusion data at 7T and the post-processing pipeline for such data. The quality of the 7T multiband data was evaluated qualitatively and quantitatively in comparison with data obtained at 3T. The acquisition of multiband two shell diffusion data allowed us to apply neurite orientation dispersion and density imaging (NODDI) to patients with glioma and to evaluate its performance in distinguishing between different types of tissue.
The results of this dissertation suggest that diffusion imaging plays an important role in assessing gliomas. These are very important steps towards improving the assessment of residual disease and distinguishing between tumor and treatment effects for patients with brain tumors.