Physiological magnetic resonance imaging methods for cancer applications

点击翻译按钮获取中文摘要

Cancer is the leading cause of death in Canada. About half of cancer patients receive conformal image-guided radiation therapy, delivering radiation to the tumour and sparing healthy tissues. Target structures are defined by manual contouring combined with atlas-based segmentation, on computed tomography (CT) images. Despite these advances, treatment has variable success. Cancer tumors feature abnormal blood supply (abnormal perfusion) and poor oxygenation (hypoxia), both markers of aggressive tumours that also affect treatment efficacy. Hypoxia takes different forms and is also a feature of other disease conditions. Disease can also decouple tissue and blood oxygenation. Physiological imaging would enable an entirely new approach to treatment targeting. We seek practical non-invasive imaging methods using magnetic resonance imaging (MRI) for (1) MR-oximetry, to map tumour oxygenation in vivo, and (2) perfusion mapping. Existing methods for oximetry, such as optical and near-infrared methods (limited to shallow tissue), polarographic (Eppendorf) electrodes (too invasive), positron emission tomography (lack of anatomical soft tissue imaging), fluorine-19 MRI (not easily accessible for human imaging), and electron spin resonance (not feasible in humans) do not provide what we seek. Instead, we explore proton R1 relaxation for tissue oxygenation, and R2* for blood oxygenation. For perfusion imaging, dynamic contrast enhanced (DCE) MRI techniques provide the highest signal efficiency and can be paired with a variety of analysis techniques. We will study fundamentals of NMR signal behaviour in tissue and develop novel techniques to address these needs. Fast, volumetric in vivo MR-oximetry will be achieved via simultaneous fat and water proton relaxation mapping. In doing so, we will investigate the relaxation mechanisms of fat-water mixtures to pursue a calibrated oximetry technique. We will advance perfusion imaging primarily using self-referenced quantitative methods, integrated with the latest multi-compartment models to account for blood flow and permeability. We will also investigate pattern recognition to draw out perfusion information from DCE-MRI, as a lean alternative for data analysis. In addition to our work in physiological imaging, we will carry on our work to generate X-ray absorption maps for radiation dose calculations from MRI data sets. This research will provide us with deeper knowledge of NMR relaxation dynamics in human tissue, and the influence of oxygen on relaxation. It will result in a novel, fast 3D MR-oximetry technique for simultaneous mapping in tissue and blood. We will know more about using MRI to estimate X-ray absorption in the body. Through this research is motivated by cancer, MR-oximetry would undoubtedly translate to a variety of applications, such as the study and characterization of metabolic diseases such as diabetes, and in exercise science, to have an impact well beyond the scope of this program.