Optimization-based optical and terahertz imaging
Abstract
Many cancer cells over-express folate receptors, and this provides an opportunity for folate-targeted fluorescence imaging and the development of targeted anti-cancer drugs. Lacking is a tool for in vivo monitoring and evaluation of the drug delivery and release kinetics in preclinical studies. Here, an in vivo deep tissue optical imaging modality that allows the study of pharmacokinetics through the disulfide bond reduction of a folate-targeting fluorophore pair is presented. The imaging method incorporates fluorescence and kinetic compartment model into optical diffusion tomography (ODT) and uses a nonlinear optimization inversion to form images of optical properties and kinetic parameters. The nonlinear inversion is also applicable to general optimization-based imaging problems, including the imaging of dielectric inhomogeneities with time-domain terahertz (THz) pulsed measurements. The significance of the THz regime lies in the accessibility of both structural and spectroscopic information. In this work, the reconstructions of real dielectric constant images are presented. Strategies for improving optimization convergence and avoiding local minima are explored, including finding the optimal regularization term and applying a dynamic cost threshold. The computational efficiency can also be enhanced by multigrid, which solves the inverse problem on a lower resolution before mapping to a higher resolution.
Degree
Ph.D.
Advisors
Webb, Purdue University.
Subject Area
Electrical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.