Light Transport in Tissue |
AcknowledgementsI would like to express my gratitude to Dr. Welch who provided funding, much needed perspective, and an introduction to the art of the absence of politics; to Dr. Valvano for the Macintosh II used for all parts of this dissertation, to Dr. Jacques for ``rolling up his sleeves and getting his hands dirty;'' to Dr. Pearce for stimulating (nay exhilarating) Friday afternoon research meetings and general harassing; to Dr. van Gemert who insisted that what I was doing was ``not unimportant;'' to Dr. Yoon for arguments about optical stuff and philosophy; and to almost doctor Cheong for always being too kind. Finally, I must thank Anne for loving me through the evolution of each chapter as we (the chapter and I) metamorphosed from a glimmer of an idea, into equations, graphs, figures and tables, into a few typeset pages, and finally into a monster. Without Anne I would not be finished yet, and may not have ever finished.
Abstract:
Two numerical solutions for radiative transport in tissue are
presented: the Monte Carlo and the adding-doubling methods.
Both methods are appropriate for tissues with internal reflection
at boundaries and anisotropic scattering patterns. The adding-doubling
method yields accurate solutions in one-dimension. The slower
Monte Carlo method is the only exact solution available for finite
beam irradiance of tissue. Convolution formulas for calculation
of fluence rates for circularly symmetric flat and Gaussian irradiances
using the Monte Carlo impulse response are presented.
The delta-Eddington method is extended to include many boundary conditions appropriate for tissue optics. The delta-Eddington method is compared with exact methods. Delta-Eddington reflection and transmission are least accurate for thin tissues and mismatched boundary conditions. Fluence calculations obtained with the delta-Eddington approximation are inaccurate (>50% error) for tissues with both mismatched boundaries and high albedos. A method and theory for the measurement of the phase function of tissue is presented. The method is shown to have a tendency to overestimate the isotropic scattering component in tissues with mismatched boundaries. A graph is presented to correct the overestimate. The backscattered peak in goniophotometer measurements is shown to result from reflection of the forward peak and not from a backward peak in the phase function. Measurements on human dermis indicate that the phase function can be described by a modified Henyey-Greenstein phase function.
A practical method for measuring the optical properties of tissue
as a function of wavelength is presented. Evaluation of the
technique indicates that the method is accurate to 10% for all
optical properties of tissue when sample thicknesses exceed one
optical depth. This technique is applied to bloodless human
dermis as a function of wavelength and to bloodless human aorta
during moderate power (
|
100mW/mm2) argon laser irradiation
as a function of irradiation time.