Robust Magnetic Resonance Imaging of Short T2 Tissues

Abstract

Tissues with short transverse relaxation times are defined as ‘short T2 tissues’, and short T2 tissues often appear dark on images generated by conventional magnetic resonance imaging techniques. Common short T2 tissues include tendons, meniscus, and cortical bone. Ultrashort Echo Time (UTE) pulse sequences can provide morphologic contrasts and quantitative maps for short T2 tissues by reducing time-of-echo to the system minimum (e.g., less than 100 us). Therefore, UTE sequences have become a powerful imaging tool for visualizing and quantifying short T2 tissues in many applications. In this work, we developed a new Flexible Ultra Short time Echo (FUSE) pulse sequence employing a total of thirteen acquisition features with adjustable parameters, including optimized radiofrequency pulses, trajectories, choice of two or three dimensions, and multiple long-T2 suppression techniques. Together with the FUSE sequence, an improved analytical density correction and an auto-deblurring algorithm were incorporated as part of a novel reconstruction pipeline for reducing imaging artifacts. Firstly, we evaluated the FUSE sequence using a phantom containing short T2 components. The results demonstrated that differing UTE acquisition methods, improving the density correction functions and improving the deblurring algorithm could reduce the various artifacts, improve the overall signal, and enhance short T2 contrast. Secondly, we applied the FUSE sequence in bovine stifle joints (similar to the human knee) for morphologic imaging and quantitative assessment. The results showed that it was feasible to use the FUSE sequence to create morphologic images that isolate signals from the various knee joint tissues and carry out comprehensive quantitative assessments, using the meniscus as a model, including the mappings of longitudinal relaxation (T1) times, quantitative magnetization transfer parameters, and effective transverse relaxation (T2*) times. Lastly, we utilized the FUSE sequence to image the human skull for evaluating its feasibility in synthetic computed tomography (CT) generation and radiation treatment planning. The results demonstrated that the radiation treatment plans created using the FUSE-based synthetic CT and traditional CT data were able to present comparable dose calculations with the dose difference of mean less than a percent. In summary, this thesis clearly demonstrated the need for the FUSE sequence and its potential for robustly imaging short T2 tissues in various applications.