论文部分内容阅读
目的:发展一种基于计算机断层扫描体系评估附件包块发展为卵巢癌的风险概率。方法:收集2011~2013年间收治的324例进行附件包块手术的患者,所有患者进行外科手术前CT检查和CA-125检测,通过统计学推导出能预测恶性肿瘤发生概率的公式,测试其灵敏度和特异性。结果:共有225例(69.4%)的患者为良性肿瘤,99例(30.6%)为恶性卵巢包块。采用逻辑回归推到恶性肿瘤发生概率可以用下面方程估计:e-3.637 2+0.030 6×(A)+0.01×(B)+0.876×(C)+1.551×(D)+1.736 77×(E)+2.67×(F)1+e-3.637 2+0.030 6×(A)+0.01×(B)+0.876×(C)+1.551×(D)+1.736 77×(E)+2.67×(F)(A:年龄,B:CA-125,C:固体包块为1,囊肿为0;D:腹水为1;E:有大网膜块为1,无为0;F:淋巴结≥1 cm为1,<1cm为0)。自然对数e为常数2.718 281 828。计算受试者工作特性曲线(ROC)下面积为81.6%。结论:该研究使用CT扫描六项参数,推导出一条能有效、灵敏预测附件包块发展为恶性肿瘤的概率的公式,协助医生优化手术方案,预期恶性病变结果,提高治疗效果。
OBJECTIVE: To develop a method based on the computed tomography system to assess the risk of adnexal masses developing into ovarian cancer. Methods: A total of 324 patients undergoing adjuvant massaging were enrolled between 2011 and 2013. All patients underwent preoperative CT examination and CA-125 test. The formula for predicting the incidence of malignant tumors was statistically obtained and their sensitivity And specificity. RESULTS: A total of 225 patients (69.4%) had benign tumors and 99 (30.6%) had malignant ovarian masses. The probability of a malignant tumor pushed by logistic regression can be estimated by the following equation: e-3.637 2 + 0.030 6 × A + 0.01 × B + 0.876 × C + 1.551 × (D) +1.736 77 × (E + 2.67 × 1 + e-3.637 2 + 0.030 6 × A + 0.01 × B + 0.876 × C + 1.551 × (D) +1.736 77 × (E) + 2.67 × (F (A: age, B: CA-125, C: solid mass is 1, cyst is 0, D: ascites is 1, E: 1, <1cm is 0). The natural logarithm e is a constant of 2.718 281 828. The area under the receiver operating characteristic curve (ROC) was calculated as 81.6%. Conclusion: This study used the six parameters of CT scan to deduce a formula that can effectively and sensitively predict the probability of adnexal masses developing into malignant tumors, to help doctors optimize the surgical plan, predict the outcome of malignant lesions and improve the therapeutic effect.