卵巢肿瘤患者年龄、CA125值及超声测量参数的Bayes分析

Bayes model of age,CA125,and ultrasonographic parameters in screening ovarian neoplasms

目的探讨Bayes分析指导卵巢肿瘤诊断的可行性及其意义。方法将卵巢肿瘤患者分为良性肿瘤组、交界性肿瘤组和恶性肿瘤组三组,选取每组患者的年龄、CA125、肿块最大径及肿瘤内动脉的阻力指数(RI)作为代表因素;比较三组间代表因素间的差异,最后行Bayes判别分析。结果三组卵巢肿瘤患者的年龄、CA-125、肿块最大径及肿瘤内动脉RI差异有统计学意义(P<0.05)。Bayes分析法判断良性肿瘤组、交界性肿瘤组和恶性肿瘤组正确率分别为76.92%、70.00%及74.51%。结论年龄、CA125、肿块大小及肿瘤内动脉RI的Bayes逐步判别分析用于指导卵巢肿瘤的诊断具有一定的可行性。

Objective To explore the feasibility and the value of Bayes analysis in clinical decision of ovari- an neoplasms. Methods A total of 152 ovarian neoplasms were divided into 3 groups：benign group（ n = 51 ） , border- line group（ n = 10） and malignant group（ n = 9t ）. We took age, serum CA125 level, ultrasonographic parameters and Doppler blood flow signals as differential diagnosis variable. Results The parameters of age, serum CA125 level, ul- trasonographic parameters in benign groups had statistical difference with those in malignant groups （ P 〈 0.05 ）. The accuracy of Bayes analysis rate was 76.92% ,70.00%, and 74.51% respectively. Conclusion Combination of sono- graphic and clinical feature might be helpful for making differential diagnosis. It is feasible to use Bayes analysis for the management of ovarian neoplasms.