| Description |
Recently, techniques from topological data analysis have been applied in various scientific domains ranging from brain imaging to sensor networks. In this thesis, we apply topological techniques in signal processing to study core body temperatures of mice. We focus on the exploration of biological events in a mouse model, in particular, pregnancies and jet lag. By combining signal processing techniques with cohomology parameterization, we create intuitive and informative visual profiles encoding periodic variations in mice temperature data. Using such visual profiles, we detect and separate successful pregnancies and pregnancy-like events. We further study the effect of jet lag on mice by clustering them into subcategories based on their topologically induced visual profiles. Jet lag has been shown to arise from critical biological factors such as travel, depression, liver cancer, and Alzheimer's disease. Our techniques have the potential to lead to new research directions in biology that investigate individual-level responses and variabilities in jet-lagged mice. |
