UC Irvine

This dissertation is an exploration of a new phase function and polarization sensitive perturbation Monte Carlo (pMC) method applied to a realistic tissue scattering model. This project is motivated by the need to improve upon previous methods that screen for early dysplastic tissue changes by optically probing suspicious regions. The majority of early dysplastic changes occur in the epithelial layer of tissue. Common strategies for increasing collection of photons that have probed shallow depths involve using short source-detector separations or using angled detector fibers. Tracking the degree of depolarization is also helpful in differentiating between photons that have probed deeper depths from others that have only probed shallow depths. Conventional Monte Carlo simulations are often employed to solve the Radiative Transport Equation (RTE) and estimate measured signals in problems with short source-detector separations or other atypical optical probe geometries.

In the pMC method, one conventional Monte Carlo simulation is run with baseline optical properties. This results in a database of photon biographies which are then subjected to pMC equations that allow for quick signal estimates to be obtained at perturbed model parameter values, bypassing the need for another conventional Monte Carlo simulation to be run. Doing this eliminates the need to do another Monte Carlo simulation to obtain reflectance estimates at slightly different optical properties. In addition to the increase in computational efficiency, this method has other advantages that may be useful to researchers interested using optical probes to identify dysplasia such as the ability to track polarization and the ability to account for changes in phase function. This is useful since polarization is employed in several optical probes designed for detection of dysplasia and the polarization state of collected photons is useful in obtaining additional information on the medium and the nature of the light-tissue interactions of collected photons. The ability to account for changes in phase function is also useful since changes in scatterer shape are associated with changes in phase function. Together, this new pMC algorithm is fine-tuned to enable other researchers to model reflectance changes for typical optical probe geometries optimized for detection of dysplasia.