Fast 4D spectral-spatial electron paramagnetic resonance imaging for in vivo oxymetry

Abstract

The present work describes our efforts thus far toward the acquisition of quantitative 3D oxygen concentration maps for tumor volumes using electron paramagnetic resonance imaging (EPRI). These maps are derived from 4D spectral-spatial images that show the spatially localized absorption spectra of an introduced molecular probe. The spectral linewidth for each spatial voxel is linearly related to the local oxygen concentration. Shown here are an image of a phantom and an in vivo image of a tumor implanted in the hind leg of a mouse. We aim to acquire the 4D data sets in 15 minutes and to produce images with sub-millimeter resolution and oxygen concentration sensitivity of several torr. Slices through the 4D spectral-spatial images are shown, as well as the respective linewidth maps given by spectral fitting. The spatial distribution and linewidths measured in the phantom image correlate well with a priori expectations. The mouse tumor image shows probe and linewidth distributions that are consistent with the mouse anatomy and physiology, including a sharp distinction between the bladder and leg/tumor signals. These results demonstrate the promise of EPRI for the measurement of in vivo oxygen concentration.