| It is well known that the presence of oxygen modulates the effectiveness of radio- and chemotherapy of cancer. Cells in hypoxic regions of tumors are more resistant to radiation and many chemotherapeutic drugs than cells that are well oxygenated. Many techniques for measuring oxygen concentration have been developed; however, most are either qualitative, invasive, or reliant on pooled data that characterize tumors generally. Electron paramagnetic resonance imaging (EPR, EPRI) uniquely enables this measurement without these shortcomings.; Using EPRI, the in vivo spatial and spectral distributions of injected, stable, free radical species are measured. These radicals, or spin probes, are designed to be sensitive to specific aspects of physiology. Interactions with endogenous paramagnetic species, changes in pH, metabolic activity, and differences in rotational freedom as a function of microviscosity or temperature cause quantitative changes in the shapes of measured resonance spectra. In particular, interaction with oxygen increases the width of resonance spectra linearly with oxygen concentration. Recent developments in the synthesis of spin probes with narrow linewidths have led to increases in the EPR signal sensitivity, oxygen sensitivity, and spatial resolution. However, additional improvements in image acquisition and reconstruction are necessary to facilitate rapid, high resolution, high sensitivity in vivo oximetric imaging.; The goal of this research has been to develop additional improvements in image quality and diagnostic ability through refinements of the image acquisition parameters and the algorithms used to reconstruct images. These improvements have facilitated in vivo oximetric imaging with spatial resolution of 1–2 millimeters, oxygen sensitivity on the order of several torr, and biologically relevant acquisition times of 20 minutes. This was achieved through the design and implementation of improved data acquisition and processing algorithms based on accurate mathematical modeling of the EPRI process. These advances were evaluated using experimental phantom data and then applied to the study of oxygen concentration in mouse tumors. The measured oxygen concentration distributions were compared to Blood Oxygen Level Dependent (BOLD) MRI images of the same mouse tumors. |
