Abstract:Based on the available data, this paper analyzes the differences of existing Venous ThromboEmbolism risk factors, and constructs the VTE risk prediction model of breast cancer patients, so as to provide certain reference for breast cancer patients to prevent and predict the risk of VTE. Based on literature analysis, the risk factors of VTE in breast cancer patients are summarized, and 10 related risk indicators of 274 cases of breast cancer patients with VTE are collected. The significant differences of corresponding indicators between VTE patients and non VTE patients are explored by nonparametric test, and the VTE risk prediction model of breast cancer patients is constructed based on logistic regression. The results show that surgical history, D-dimer level, lipoprotein, plasma protein C activity and C-reactive protein are important risk factors for VTE in breast cancer patients. Based on these risk factors, the logistic regression model is constructed, and the average prediction accuracy of ten fold cross validation is 75.36%.