Quantitative Perfusion-Sensitive Mri Phantoms
Perfusion-sensitive MR methods are increasingly utilized in preclinical and clinical MR research studies with the promise of providing quantitative estimates of parameters that describe in vivo microvasculature. One of these techniques, dynamic contrast enhanced (DCE) MRI, has found particularly common use in oncology for the detection, staging, and monitoring of highly vascularized tumors. DCE-MRI has been qualitatively validated by various studies that show a high correlation between modeled
... on between modeled parameters from DCE and histologically measured microvascular density (MVD). However, in the absence of a matching "gold-standard" technique, DCE-MRI has not yet been quantitatively validated (i.e., the accuracy of the estimated parameters is unknown). Partly because of this inability to determine the accuracy of the measured parameters, there remains debate in the literature about which DCE signal model (s) best reflect(s) experimental data. In order to address these scientific challenges, realistic DCE tissue phantoms have been constructed. These phantoms implement semi-permeable hollow fibers, found commonly in commercial hemodialysis cartridges, to simulate "leaky" vasculature. Their design and construction are cataloged in this thesis. In addition, the phantoms have been experimentally characterized. In conjunction with these experiments, an interesting example of diffusion driven longitudinal relaxation was observed and is described herein. Lastly, the permeability of the fiber wall with respect to Gd-based contrast agents has been measured independently and compared with values derived from a mock-DCE experiment performed on the phantoms. In general, the results of these experiments support current DCE-MRI methods. iii ACKNOWLEDGMENTS For the completion of this major thesis project I would like to thank my wife, children, and parents for their support and inspiration. I thank my advisors Joe Ackerman and Joel Garbow for their time, talents, insights, corrections, endless editing, and patience. I am grateful to them for taking interest in my research in the midst of many other responsibilities and commitments. Without their expertise and guidance this research would not have been possible. I thank John Engelbach and Kassie Chaffee for their friendship and help. I thank others who have specifically contributed to publications resulting from this research: Jeffrey Arbeit, X.