影响腮腺癌术后无瘤生存期电子计算机多因素分析  

A Multivariate Regression Analysis with Computer on the Prognosis of Parotid Carcinoma after Operation

作　　者：杨斌[1] 张永福[1] 王庆延[2] 俞俊甫[2] 

机构地区：[1]江西医学院口腔医学系 [2]江西医学院计算机中心

出　　处：《江西医学院学报》1991年第1期32-35,共4页Acta Academiae Medicinae Jiangxi

摘　　要：本文以43例临床和随访资料完整的腮腺癌手术治疗病人为背景,选出包括生化和免疫指标在内的13个项目、19个因素,应用电子计算机多角度综合分析腮腺癌预后影响因素。建立预测腮腺癌术后无瘤生存期的数学模型;筛选出影响无瘤生存期的七项重要因素:病期、癌分化程度、面瘫、淋巴结转移、手术方式、癌细胞生长指数和癌中巨噬细胞浸润密度。对所建模型考核证实:多元回归在α=0.01水平上显著,其判断准确率为83％,敏感度84％,特异度83％。Forty-three patients after surgical treatment of parotid carcinoma with adequate records , both clinical and follow-up were selected for study. Basing on multivariate analysis with IBM personal computer, a mathematical model was developed from 13 items and 19 factors (including biochemical and immunological criteria) to predict the disease-free interval after operation. According to the result of mutivariate analysis, there were 7 important factors influencing the postoperative disease-free interval of the patients with parotid carcinoma. They are the course of disease ,degree of differentiation, paralysis of facial nerves, lymphnode metastasis,method of operation, proliferative index of cancer cells and density of infiltrating macrophages. The results proved that:The multivariate regression model was significant at a=0.01 level. Its accurate rate was 83% ,sensitivity 84% and specificity 83%.

关 键 词：腮腺肿瘤 预后 诊断 计算机辅助 模拟-数字转换 

分 类 号：R739.870.7[医药卫生—肿瘤]