Font Size: a A A

A Study On Life Cycle And Influencing Factors Of Breast Cancer

Posted on:2020-04-15Degree:MasterType:Thesis
Country:ChinaCandidate:S J XieFull Text:PDF
GTID:2370330623960340Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
Breast cancer is the most common female malignant tumor occurring in the epithelial tissue of the breast.Global incidence of breast cancer has been on the rise since the late1970s,developed countries like America and Europe always had a high incidence of breast cancer[29],especially in the economically developed cities,the incidence of breast cancer is especially high.In this paper,in order to explore the pathogenic factors and treatment methods of breast cancer,data from Pereira B.[12]in the journal Nature Communications were analyzed.Kaplan-meier estimation was performed[8],describing ER status,HER2 status,chemotherapy,hormone therapy,surgical type and overall survival curve.First in this paper,through the establishment of risk of Cox proportional regression model and accelerate the failure model,indicates the Nottingham prognostic index,age,chemotherapy,hormone therapy and surgical type have a significant impact on the survival of patients with breast cancer,the bigger Nottingham index and age,the shorter the survival time,the bigger the chemotherapy,hormone therapy and breast surgery patients survival probability;Then the survival state of patients was predicted by constructing decision tree model and random forest model.By comparing Cox proportional risk regression model[36]and random forest model[53][54],it was found that Cox model was more suitable to explain the influence of covariates on survival time,and random forest model had better effect in predicting survival cycle.Finally,a quantile regression model was established to analyze the influence of covariates when the survival rates were different.The concept of early detection and early treatment is proposed,and chemotherapy,hormone therapy and breast conserving surgery are adopted as far as possible to increase the survival probability of breast cancer in China,promoting the scientific cognition of breast cancer in our country and helping medical personnel to make the best treatment plan.
Keywords/Search Tags:Breast Cancer, Censoring Data, K-M Estimator, Cox hazard proportional regression model, Accelerated failure model, Random forest mode, Quantile regression
PDF Full Text Request
Related items