Scalability of time-domain electron paramagnetic resonance in vivo imaging

The research in this dissertation presents the work done in enhancing the scalability of time-domain electron paramagnetic resonance (EPR) imaging, and specifically the electron spin echo (ESE) technique in several aspects. First, the issue of imaging frequency bandwidth is addressed in the context of high-resolution imaging, where the multi-stepped Zeeman offset methodology is provided as a means to extend the effective bandwidth. High spatial resolution ESE images are obtained using large applied magnetic field gradients. Next, the same multi-field bandwidth-extension technique is used to acquire ESE images from larger objects. We show that it is possible to image objects comparable to human tumors in size, without sacrificing image resolution. In the third chapter, which also deals with large object imaging, the bandwidth extension technique is enhanced one step further: Based on the prior information about the object size and boundaries, we match the size of acquisition bandwidth to the size of each single projection. This allows further enhancing of the image signal-to-noise ratio (SNR) and reduction of image artifacts. Scaling of ESE imaging in speed requires use of high repetition rate for the imaging pulse sequence. We discovered that when the pulse repetition time is comparable to the spin-lattice relaxation time, a bias in estimation of spin-spin relaxation time, and subsequently the oxygen tension estimate is introduced. By employing dynamically adjusted repetition rate schemes, we have shown that the saturation-based bias in our estimate can be minimized with some modest reduction in acquisition SNR rate. In sum, methodologies for scaling of ESE in resolution, size and speed are shown to be effective in acquiring of oxygen images.