Polarization-based optical imaging and processing techniques with application to the cancer diagnostics

In this investigation, a polarization-based imaging system is developed and described that measures the two-dimensional effective backscattering Mueller matrix of a sample in near real-time. As is well known, a Mueller matrix can provide considerable information on the makeup and optical characteristics of a sample and also directly describes how the sample transforms an incident light beam. The ability to measure the two-dimensional Mueller matrix of a biological sample, therefore, can provide considerable information on the sample composition as well as the potential to reveal significant structural information that normally would not be visible through standard imaging techniques. Additional information can also be obtained through the application of image-processing, decomposition, and reconstruction techniques that operate directly on the 2D Mueller matrix. Using the developed system, it is shown how the induction of internal strain within the sample coupled with image reconstruction and decomposition techniques can further improve image contrast and aid in the detection of boundaries between tissues of different biomechanical and structural properties. The studies presented were performed with both rat tissue and a melanoma-based tissue culture. The results demonstrate how these techniques could provide information that may be of diagnostic value in the physical detection of malignant lesion boundaries.