| Popis: |
Cancer is the second most common cause of death in the US and is responsible for about 25% of the deaths according to the American Cancer Society. There is a need for quick and accurate detection of tumors in a real time automated environment without the delays associated with laboratory procedures. There is also a need to identify the exact location and extent of the tumor tissues so that surgeries would not only be more effective but also as minimally invasive as possible. The purpose of this study was to distinguish the margins between healthy and cancerous tissues on slides containing samples of mice brain tumor through the use of polarimetric stokes imaging techniques. Comparison between the cancerous and healthy tissue portions from polarimetric imaging was expected to provide better distinction of the two areas and the margin between them.The experiments were conducted with a broadband white light source and the data was recorded using a circular polarized rotating retarder system in transmission mode. The samples used were unstained sliced sections of mice brain prepared using standard procedures. The images were saved with an optical camera that was impartial to color or fluorescence properties of the samples. The Stokes parameter images of the samples were computed through the Fourier analysis method. The images for Degree of Polarization (DOP), Degree of Linear Polarization (DOLP) and Degree of Circular Polarization (DOCP) were also computed. The images thus generated were segmented and post-processed to analyze properties such as texture and tissue density. Additionally, small selections of tumor and healthy tissue having the same pixel area were chosen from the Stokes images for further statistical analyses.The statistical findings from this study showed that image features of tumor tissue had lower variance values to the various degrees of polarization when compared to that of normal tissue. This meant that the surface of tumors were more densely packed and more uniform than healthy tissue thus enabling successful segmentation, thereby proving the expectation that polarimetric stokes imaging was an efficient and robust tool in the detection of margins. |
