This thesis describes the development of a combined label-free imaging and analytical strategy for intraoperative characterization of cancer lesions using the coherent anti-Stokes Raman scattering imaging (CARS) technique. A cell morphology-based analytical platform is developed to characterize CARS images and, hence, provide diagnostic information using disease-related pathology features. This strategy is validated for three different applications, including margin detection for radical prostatectomy, differential diagnosis of lung cancer, as well as detection and differentiation of breast cancer subtypes for in situ analysis of margin status during lumpectomy. As the major contribution of this thesis, the developed analytical strategy shows high accuracy and specificity for all three diseases and thus has introduced the CARS imaging technique into the field of human cancer diagnosis, which holds substantial potential for clinical translations. In addition, I have contributed a project aimed at miniaturizing the CARS imaging device into a microendoscope setup through a fiber-delivery strategy. A four-wave-mixing (FWM) background signal, which is caused by simultaneous delivery of the two CARS-generating excitation laser beams, is initially identified. A polarization-based strategy is then introduced and tested for suppression of this FWM noise. The approach shows effective suppression of the FWM signal, both on microscopic and prototype endoscopic setups, indicating the potential of developing a novel microendoscope with a compatible size for clinical use. These positive results show promise for the development of an all-fiber-based, label-free imaging and analytical platform for minimally invasive detection and diagnosis of cancers during surgery or surgical-biopsy, thus improving surgical outcomes and reducing patients' suffering.
| Identifer | oai:union.ndltd.org:RICE/oai:scholarship.rice.edu:1911/70246 |
| Date | January 2012 |
| Contributors | Wong, Stephen T.C. |
| Source Sets | Rice University |
| Language | English |
| Detected Language | English |
| Type | Thesis, Text |
| Format | 111 p., application/pdf |
Page generated in 0.0023 seconds