Advancing MR-based elastography through improved instrumentation, resolution, and materials modeling

The focus of this Dissertation research is to develop a noninvasive imaging tool that provides detailed information about the viscoelastic material properties of biological tissues at the microscopic scale. Magnetic Resonance Elastography (MRE) creates contrast based upon tissue's viscoelastic properties. Microscopic-MRE (microMRE) could be used to detect cancer in its very early stages or assess the structural integrity of engineered tissue specimens as they differentiate. With the goal of improving the resolution and accuracy of microMRE, the research objectives undertaken as part of this Dissertation study have been to improve microMRE: (1) instrumentation, (2) reconstruction algorithms, and (3) tissue viscoelasticity models used in MRE reconstruction.;Instrumentation. Improving microMRE spatial resolution requires new techniques for creating vibratory shear waves at higher frequencies with shorter wavelengths. Two novel approaches are investigated: (i) using a needle-type actuator driven axially within the specimen to generate cylindrical shear waves; and (ii) using modulated radiation force of ultrasound to remotely create a focused shear wave source within the region of interest. The design of these actuators is optimized via a combination of experimental and computational studies.;Reconstruction. The resulting vibro-acoustic shear wave propagation is imaged via a high field phase contrast MR at particular frequencies. Based on this information it is then desirable to locally estimate the tissue material properties, such as elasticity and viscosity. Numerous techniques have been applied in the past (for macroscopic resolution MRE) ranging from simple measurements of wavelength and attenuation, to more sophisticated finite element (FE) model approaches. In the present work, a FE-based approach is improved and tailored to the microscopic scale.;Tissue viscoelasticity modeling. Reconstruction can proceed assuming standard viscoelastic tissue constitutive relations, such as a Voigt or Maxwell model. However studies have shown that these models are limited in accuracy. Recently, fractional order material models have shown promise in tissue mechanics modeling. Their application in microMRE-based reconstruction of tissue phantom properties is explored. In order to validate the microMRE material estimates and compare the accuracy of different material models three benchmark examples of shear dominant waves are studied.