Cryosurgical monitoring using visible light imaging and electrical impedance tomography: A feasibility study

Cryosurgery, which uses low temperatures to destroy unwanted tissue, relies heavily on medical imaging to monitor the freezing process. While existing imaging techniques are adequate in some cases, many procedures would benefit from alternative monitoring methods. Two promising techniques that have not been previously assessed for cryosurgical use are optical imaging (OI) and electrical impedance tomography (EIT). OI images are based on differences in visible radiation attenuation, while EIT images are maps of electrical impedance distribution. The act of freezing tissue increases both its optical attenuation and electrical impedance. In this dissertation, we evaluate the feasibility of using OI and EIT to monitor cryosurgical procedures.; In order to test whether frozen tissue provides sufficient optical contrast for OI, we designed an experimental chamber capable of measuring transmitted and reflected visible light through an MRI-monitored bulk tissue sample undergoing one-dimensional planar freezing. We found that the increased scattering of the frozen region attenuated the optical signal in both reflection and transmission modes. The results of a diffusion-based finite difference model analogous to the experimental system compared favorably to our experimental data.; A similar approach was used to test whether frozen tissue provides adequate electrical impedance contrast for EIT. A small planar tissue sample within an impedance measurement chamber was one-dimensionally frozen using a directional freezing stage. The highly insulative frozen region produced easily discernible impedance changes as the ice front moved across the sample. Results produced by a finite element model of our sample chamber compared well to all measured samples.; In the final study, we developed an image reconstruction algorithm appropriate for cryosurgery. This boundary element-based method adjusts the geometry of the imaged frozen region rather than the impedances of the individual pixels, as is the case with traditional finite element-based algorithms. This Front-Tracking (FT) method takes advantage of the fact that frozen regions in cryosurgery are generally simply shaped, convex, have known origins, and are of known resistivity. Simulations based on identical phantoms revealed that our FT algorithm requires fewer independent voltage measurements, achieves finer edge resolution, and is computationally more efficient than traditional EIT reconstruction algorithms.