Stefan Elrington

Magnetic resonance imaging (MRI) is the leading non-invasive imaging technique of soft tissues on anatomical and millimeter scales. In conventional MRI, hydrogen-1 in water and liquid fats are detected; the relatively narrow spectra of these signals are the key feature that enables MR imaging with high-spatial resolution (i.e., sub-mm in each dimension). The broad NMR spectra of solids would be much more difficult to use in MRI, and would ordinarily result in a low spatial resolution image. Previously, our lab developed a pulse sequence to effectively narrow the broad spectrum of lines in solids. This sequence, applied to phosphorus-31 spins in bone, was able to achieve high resolution imaging. Despite this tremendous progress, our approach to MR imaging of solid samples is still signal limited and the imaging times are very long. In this thesis work, I propose using a variation on a solid-state, double-resonance NMR technique to increase the signal of phosphorus-31 imaging and to be able to do so in a shorter time. This approach uses cross-polarization to resonantly transfer magnetization from a hydrogen-1 spin bath that is cooler, i.e. more polarized, and more quickly refreshed, to a phosphorus-31 spin bath. As explained in this thesis work, bone is a poor choice of sample for a cross-polarization experiment.  In order to overcome this disadvantage, I developed a new approach, called StepCP, to boost the phosphorus-31 signal in a stepwise fashion. Using this technique, I demonstrate that the signal to noise ratio using double-resonance in solids can be increased, and that the signal can be attained in fraction of the acquisition times previously used to generate phosphorus-31 images in bone.