Search

Quantitative tissue optical spectroscopy has been considered as a promising method for clinical diagnosis, owing to its ability to non-invasively give an objective assessment of biological tissues at cellular and sub-cellular levels. In spite of recent advances in optics and the computational power, not many quantitative tissue optical sensing technologies have been translated into clinical practice. In order to translate this technology in the clinics, we need to further improve the technology. To name a few, we need accurate and rapid quantification method for a real-time diagnostic feedback. Next, we need computational methods for complex tissue-optics problems. Also, we need a novel approach in probe design for the inaccessible organs. This dissertation focuses on the development, verification and validation of theoretical (mathematical and computational) and experimental (instrumental) tool set to promote the translation of quantitative tissue optical spectroscopy into clinical diagnostic applications. For the mathematical tool, a direct-fit photon tissue interaction (DF-PTI) model that could rapidly and accurately extract the parameters associated biophysical features was developed and validated to characterize the precursor lesions of pancreatic cancer. A rapid scattering model on pancreatic tissue reflectance based on principal components analysis (PCA) results was also developed. The diagnostic capability of scattering properties obtained was demonstrated on an 18-patient data set using a rigorous statistical method, which implied the potential of reflectance spectroscopy for real-time detection of pancreatic cancer. For the computational tool, a ray-traced Monte Carlo (RTMC) simulation for the design of fluorescence spectroscopy or imaging system utilizing complex optics to probe turbid biological tissues was devised. This new method was verified computationally with epithelial tissue models and experimentally using tissue-simulating optical phantoms. For the instrumental tool, the design and development of minimally-invasive diagnostic technologies employing optoelectronic components were discussed. In this dissertation, we focused on detection of pancreatic cancer, which has the worst prognosis among other major cancers. Pancreatic tissues were employed as our model system to validate our developed tools. The developed technology and tools can be applied to a variety of other human tissue sites to help in the translation of quantitative tissue optical sensing in a clinical setting.