Simultaneous Multislice Functional Magnetic Resonance Imaging.

magnetic resonance imaging;functional magnetic resonance imaging;simultaneous multislice;multiband;coil compression;Biomedical Engineering;Electrical Engineering;Neurosciences;Psychiatry;Radiology;Engineering;Health Sciences;Science;Biomedical Engineering

Chu, AlanPeltier, Scott J ;

University of Michigan

关键词: magnetic resonance imaging;    functional magnetic resonance imaging;    simultaneous multislice;    multiband;    coil compression;    Biomedical Engineering;    Electrical Engineering;    Neurosciences;    Psychiatry;    Radiology;    Engineering;    Health Sciences;    Science;    Biomedical Engineering;   

Others  :  https://deepblue.lib.umich.edu/bitstream/handle/2027.42/120669/alanchu_1.pdf?sequence=1&isAllowed=y

瑞士|英语

来源: The Illinois Digital Environment for Access to Learning and Scholarship

【 摘 要 】

Functional magnetic resonance imaging (fMRI) is a valuable tool for mapping brain activity in many fields. Since functional activity is determined by temporal signal changes, undesired fluctuations from physiological motion are problematic. Simultaneous multislice (SMS) imaging can alleviate these issues by accelerating image acquisition, increasing the temporal resolution. Furthermore, some applications require a temporal resolution higher than what conventional fMRI will allow. Current research in SMS has focused on Cartesian readouts due to their ease in analysis and reconstruction. However, non-Cartesian readouts such as spirals have shorter readout times and better signal recovery.This work explores the acquisition and reconstruction of both spiral and concentric ring readouts in parallel SMS. The concentric ring readout retains most of the benefits of spirals, but also increases the usability of alternative reconstruction techniques for non-Cartesian SMS such as generalized autocalibrating partially parallel acquisitions (GRAPPA). To date, non-Cartesian SMS imaging has only been reconstructed with sensitivity encoding (SENSE), but results in this work indicate GRAPPA-based reconstructions have reduced root-mean-square-error compared to SENSE and good subjective image quality as well. Furthermore, using point spread function analysis, the concentric ring trajectory is found to have superior slice separation properties compared to a spiral one.Since parallel imaging greatly magnifies the amount of data used for reconstruction, a novel coil compression method is developed, which outperforms conventional coil compression in fMRI, substantially decreasing the amount of reconstruction time needed for sufficient detection of functional activation. Results indicate that the proposed method can compress 3 simultaneous slice data using a 32-channel coil down to only 10 virtual coils without any adverse effects in functional activation, noise, or image artifacts. Competing methods require substantially more coils for preservation of the data, resulting in large reconstruction time savings for the proposed method.This work also explores the use of Hadamard-encoded fMRI for increased temporal resolution. Because Hadamard-encoded SMS uses data from multiple time frames to separate slices, physiological noise correction is critical. However, even with physiological noise correction, results indicate Hadamard-encoded fMRI is not as reliable as conventional fMRI due to undesired temporal fluctuations, most notably from uncorrected physiological noise.