| Summary: | 碩士 === 國立中山大學 === 應用數學系研究所 === 102 === Since 1982, cancer has become one of the leading causes of death in Taiwan. The focus of this thesis survival analysis for ovarian cancer.
The first part of the thesis is analysis to the risk factors for recurrence in patients with borderline ovarian tumors. We used Cox proportional hazard model to analyze the risk of many factors.
In the second part, we analyzed the risk factors for recurrence in patients with invasive ovarian tumors. These patients have longitudinal CA125 data. So we employed a joint model to explore the relationship between longitudinal CA125 profile and recurrence. In the final part of this section, we predicted patients’s recurrence probabilities after three years post surgery to validate our model.
In the last part, we estimated the time-varying effects of the risk factors in ovarian cancer. In the Cox proportional hazards model, the hazard ratio is assumed to be indepen- dent of time. But in fact, the hazard ratio of many variables is not a constant. So applied an extended Cox model to estimate how the hazard ratio of these time-varying coefficients varies with time.
|
|---|
