| With the rapid development and widespread application of the bioimaging technologies, we are entering a new era of bioimaging sciences, in which we are becoming able to model, measure and monitor biological processes at multiple levels. By virtue of its ability to image the structure, metabolism, and function of internal tissues/organs of humans or any biological objects, magnetic resonance imaging (MRI) has played and will continue to play a significant role in biology and medicine in the years to come. A well-known limitation of MRI is its relatively low imaging speeds, significantly hindering its applications. To address this fundamental problem, many fast imaging methods have been proposed for the past three decades. A recent exciting progress in this field is the emergence of the parallel MRI technology, which promises to provide another powerful tool to significantly accelerate the existing imaging techniques.;The dissertation seeks to establish a firm theoretical and algorithmic framework for parallel MRI to enable this promising technology to achieve maximal speed enhancements for various practical applications. Within this framework, several novel approaches are taken to address the optimal data acquisition and image reconstruction problem using sampling theory and statistical detection and estimation theory. A pulse sequence program is developed for the data acquisition scheme in the GE MRI platform, and software is developed for the image reconstruction algorithm. The research also systematically evaluates the performance of the novel approach using both phantoms and in vivo data. The in vivo data is acquired by the GE 3T MRI scanner located at Medical College of Wisconsin, with application focused on functional neuroimaging where high imaging speed is desired. The research is expected to achieve maximum imaging speed enhancements, thus producing significant results of both theoretical and practical values to the field of biomedical imaging. |
