At present, magnetic resonance imaging (MRI) does not provide the robust information required to study the malignant processes in lesions. This can make tumors difficult to diagnose and also limits the usefulness of MRI in predicting and evaluating cancer therapy. Spectroscopic images offer the ability to provide important additional biological information which can complement the water density and relaxation time data measured by conventional methods. However, a number of factors significantly limit the usefulness of such metabolite images, including low signal-to-noise ratio (SNR), poor spatial resolution, long imaging times, sensitivity to magnetic field inhomogeneities, and the inability to view important metabolites, such as lactate, in the presence of the relatively large lipid signals. We have addressed these limitations using novel pulse sequences for data acquisition and estimation theory for data reconstruction and processing. The goal of the proposed work is to further develop and validate these techniques using improved quantification of in vivo spectra, fast spectroscopic imaging methods, optimized lactate imaging pulse sequences, motion insensitivity, and image processing algorithms for resolution enhancement. These will be tested using anecdotal patient studies. Using a mouse model, we will also test the hypothesis that lactate levels are directly correlated with tumor hypoxia and, as such, may be predictive of tumor response to therapy.
Showing the most recent 10 out of 31 publications