The goal of this project is to develop new technical approaches to mitigate the effects of severely inhomogeneous and patient-dependent RF transmission (B+) fields that limit ultra-high field MRI, with particular emphasis on spin 1 echo and fast spin echo (FSE) neuroimaging at 7 Tesla (T). Our approaches comprise new techniques for patient- tailored single-channel and parallel excitation, including trajectory designs and pulse design algorithms, in spin echo acquisitions of high importance at 7 T. These methods will be essential to realizing the improvements in SNR and resolution, and exploiting the distinct contrast mechanisms that 7 T MRI promises. They stand to enhance the quality of spin echo MRI at lower clinical field strengths, and have broad impact in practical applications. Three imaging methods that stand to benefit greatly from 7 T are FSE acquisitions such as fluid-attenuated inversion recovery (FLAIR), diffusion tensor imaging (DTI), and blood oxygenation level-dependent (BOLD) functional MRI (fMRI). FSE sequences are key for imaging pathologies of the central nervous system, while DTI and BOLD fMRI have revolutionized structural and functional connectivity studies. However, at low field these sequences are hampered by limited SNR and spatial resolution (FSE and DTI), and intravascular contamination (BOLD). At 7 T, FSE and DTI data can be acquired with higher SNR and spatial resolution, and Hahn spin echo (HSE) BOLD fMRI can be used to obtain functional signals that can much more accurately localize neural activation. Unfortunately, because patient-dependent B+ field inhomogeneity causes flip angle inhomogeneity, 1 the quality of spin echo acquisitions at 7 T is currently severely limited. Recently, patient-tailored single-channel and parallel excitation methods have been developed to mitigate the effect of B+ inhomogeneity, and the basic 1 hardware and software required by these methods is becoming standard on 7 T scanners. However, there has been little development of these methods for the large-tip-angle excitations required by spin echo acquisitions. This project will address this gap in development. We will develop new three-dimensional excitation k-space trajectories for large-tip-angle slice-selective excitation, and new pulse design algorithms to design excitation and refocusing pulses using those trajectories. We will improve upon the current dominant 'spokes'trajectory, which is generally unsuitable for large-tip-angle slice-selective excitations due to the high power of spokes RF pulses. We hypothesize that alternative trajectories exist that can provide improved spatial encoding capabilities with higher spectral bandwidths that are minimally impacted by reshaping to reduce RF power deposition, and with improved flexibility in selecting slice thickness and sharpness. Our pulse design methods will enable the joint design of excitation and refocusing pulses to meet the unique demands of spin echo imaging sequences, and will not only improve the pulses'performance, but will also add to their capabilities. We will evaluate the performance of our patient-tailored pulses in FSE, DTI and HSE BOLD fMRI acquisitions, using data and image quality metrics specific to those modalities.
Spin echo magnetic resonance imaging (MRI) at 7 Tesla field strength has the potential to deliver much clearer images of brain structure and function than is provided by lower field strength MRI scanners in wide use today, which would greatly improve our understanding of the brain and the diagnostic power of brain MRI. However, spin echo image quality is severely degraded at 7 Tesla by spatial nonuniformity of the radiofrequency electromagnetic fields used in image formation. In this project we will develop methods to alleviate this problem, and validate them in spin echo neuroimaging protocols of broad interest to researchers and clinicians.
|Harkins, Kevin D; Does, Mark D; Grissom, William A (2014) Iterative method for predistortion of MRI gradient waveforms. IEEE Trans Med Imaging 33:1641-7|