| Collagen I density, distribution, and organization influence cell behavior and are potential biomarkers in a wide array of diseases ranging from osteogenesis imperfecta (OI) and muscular dystrophy, to ovarian and colon cancer. Alteration of collagen I is associated with numerous disease states including dermal wounds, vascular disease, liver fibrosis, prostate cancer, ovarian cancer, breast cancer, and asthma. For example, in breast cancer, there is clear evidence that collagen architecture plays a key role in invasion and tumor progression.;Many recent, novel characterizations elucidating the relationship between collagen and disease state have arisen due to rapid development in the area of non-invasive high-resolution imaging methods, such as second harmonic generation (SHG) microscopy, that can track spatial and temporal changes in the collagen matrix. However, even with these great advances in our knowledge and ability to image collagen, there is still a dearth of robust computational methods for quantitative analysis of collagen architecture.;Demonstrated in this dissertation are analyses of collagen architecture using the curvelet transform and intensity-based methods, and the presentation of two new software tools, CurveAlign and TumorTrace, for quantifying collagen architecture and related changes in cellular morphology and behavior. While we primarily focus on analysis of signals from SHG microscopy in mammary carcinoma, the analysis is widely applicable. |
