Computational modeling of time-resolved fluorescence transport in turbid media for non-invasive clinical diagnostics.

Fluorescence spectroscopy and imaging methods, including fluorescence lifetime sensing, are being developed for a variety of non-invasive clinical diagnostic procedures, including applications to early cancer diagnosis. Here, both the theoretical developments and experimental validations of a versatile, numerical Monte Carlo code that models photon migration in turbid media to include simulations of time-resolved fluorescence transport are presented. The developed numerical model was used to study, for the first time, the dependence of time-resolved fluorescence signals emanating from turbid media on the optical transport coefficients, fluorophore properties and source-detector configurations in single-layered turbid media as well as more complex multi-layered turbid media. The numerical codes presented here can be adapted to model a wide range of experimental techniques measuring the optical responses of biological tissues to laser irradiation and are demonstrated here for two specific applications (a) to model time-resolved fluorescence dynamics in human colon tissues and (b) to extract the frequency-dependent optical responses of a model adult human head to an incident laser-source whose intensity was harmonically modulated i.e. simulating frequency-domain measurements. Specifically, measurements of time-resolved fluorescence decays from a previous clinical study aimed toward detecting differences in tissue pathologies in patients undergoing gastro-intestinal endoscopy were simulated using the Monte Carlo model and results demonstrated that variations in tissue optical transport coefficients (absorption and scattering) alone could not account for the fluorescence decay differences detected between tissue pathologies in vivo. However, variations in fluorescence decay time as large as those detected clinically between normal and pre-malignant tissues (of 2 ns) could be accounted for by simulated variations in tissue morphology or biochemistry while intrinsic fluorophore lifetimes were held constant. Potential applications of the numerical code for the construction of optimized fiber-probes for efficient clinical diagnostics and the reconstruction of tissue optical properties to match experimental measurements, possibly in real-time via the use of heuristic scaling procedures, are discussed.Ph.D.Applied SciencesBiomedical engineeringOpticsPure SciencesUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/125246/2/3186780.pd