Improved magnetic resonance temperature imaging through pulse sequence and reconstruction techniques

This dissertation focuses on improving magnetic resonance (MR) thermometry techniques. The goal is to develop and improve data acquisition and image reconstruction methods such that MR can be used to effectively monitor a variety of thermal therapy procedures, with high intensity focused ultrasound (HIFU) treatments being of particular interest. These treatments selectively heat a small volume of tissue rapidly, making the temperature monitoring aspect both vital and challenging. Magnetic resonance imaging can measure and map temperature changes in tissue based on the temperature dependence of the resonant frequency of water protons. The first set of results given quantifies the effects that different spatial sampling properties have on these MR temperature maps during HIFU treatments. The study indicates that accurate measurement of a single point HIFU induced temperature distribution requires a spatial resolution of 1x1x3 mm or less. Covering the entire volume of interest at this resolution would take an excessively long time to acquire using conventional MR thermometry means. The remainder of the work presented in this dissertation concerns two reconstruction techniques that allow MRI temperature maps to be acquired at an accelerated rate. The first method is known as temporally constrained reconstruction (TCR). This algorithm generates images from undersampled data by iteratively minimizing a cost function.