MRI parallel imaging offers significant advantages over conventional techniques: it allows faster acquisition and reduces artifacts. The sensitivity encoding scheme (SENSE) can be applied to high-speed, single-shot MRI (such as spiral or EPI) in which speed can be traded for higher resolution or reduced off-resonance effects. Parallel imaging, when properly applied to high-speed imaging, has the potential of improving many new and evolving functional neuroimaging methodologies such as diffusion- and perfusion-weighted imaging (DWI, PWI), diffusion tensor imaging (DTI) and functional BOLD imaging (fMRI). This project will improve SENSE acquisition and reconstruction methods for both EPI- and spiral-based sequences. The overall goal is to demonstrate that SENSE can be combined with DWI/DTI to significantly improve the evaluation of patients with signs or symptoms of cerebral ischemia. These efforts are directed in specific aims; (i) to optimize and evaluate an improved Cartesian SENSE approach for high quality DWI and PWI, (ii) to optimize and evaluate self-calibrating spiral imaging using generalized SENSE (GSENSE). Specifically, we will, in three aims: (1) adapt conventional Cartesian SENSE approaches with several new hypotheses to improve DWI/DTI and PWI. To do this, a variety of sophisticated numerical techniques, including Tikhonov regularization, total least squares (TLS), and the consideration of an accurate spatial response function (SRF), will be combined to improve Cartesian SENSE reconstruction. (2) We will improve and optimize the GSENSE reconstruction by evaluating different preconditioning methods for the conjugate gradient (CG) algorithm and evaluate hybrid approaches that combine both gridding-based and faster convolution-based reconstruction techniques in order to diminish the computational demands for reconstruction. Moreover, the GSENSE improvements will focus on optimization of spiral DTI sequences for ideal variable-density (VD) spiral DTI as well as for a dual-echo-interleaved spiral out-in sequence (DEISOI). Finally, (3), we will determine the benefit from improved DWI/DTI and PWI for stroke imaging by applying our improvements in a consecutive series of 70 stroke patients over two years. The advanced SENSE methods developed in this effort will significantly improve multi-coil parallel imaging acquisition and reconstruction methods. We will demonstrate that these improvements will lead to better diffusion and perfusion imaging quantification tools that will ultimately aid clinicians in making better therapeutic decisions without the need for additional testing. Although the primary target of the research proposal is improving stroke imaging, we fully anticipate that other fields, including fMRI, cardiac/abdominal MR/, and CE-MRA, could strongly benefit from the results.
Showing the most recent 10 out of 86 publications
