论文部分内容阅读
目的:探讨彩色多普勒超声结合多变量Logistic回归法预测附件区恶性肿瘤的临床价值。方法:回顾性分析附件肿物患者110例,其中卵巢良性肿瘤49例,炎症包块11例,卵巢恶性肿瘤45例,交界性肿瘤2例,输卵管肉瘤1例,间皮肉瘤1例,伯基特淋巴瘤1例。二维超声观察肿块声像图表现并测量,然后观察子宫、附件及瘤体的血流情况,肿块血流来源,与病灶内部及周边多点取样进行多普勒频谱曲线测量,并记录阻力指数(R I)、搏动指数(PI)、收缩期峰值血流速度(PSV)、时间平均最高流速(TAMXV)等,采用多变量Logistic回归模型进行分析推测附件肿瘤为恶性的可能性。结果:多变量Logistic回归模型与进入回归方程的单一变量(R I,肿瘤形态评分,肿瘤血流位置评分,腹水)相比能更准确的预测附件区肿瘤的恶性度,差异有显著性(P<0.000 1)。预测卵巢恶性肿瘤的概率,提高其敏感度、特异度及阳性预测值,诊断概率大于25%时,对恶性肿瘤诊断的敏感度和特异度、阳性预测值及准确性最高。结论:多变量Logistic回归法诊断附件区良恶性肿瘤的诊断符合率较高,多变量Logistic回归法诊断附件恶性肿瘤的准确性比单独应用灰阶超声或者彩色多普勒超声更高,更具有诊断价值。
Objective: To investigate the clinical value of color Doppler ultrasound combined with multivariate Logistic regression in predicting the malignancy in the annex area. Methods: A total of 110 cases of accessory tumor were retrospectively analyzed. Among them, 49 cases were ovarian benign tumor, 11 cases were inflammatory mass, 45 cases were ovarian malignant tumor, 2 cases were borderline tumor, 1 case was tubal sarcoma, 1 case was mesothelioma, 1 case of special lymphoma. Two-dimensional ultrasonography was performed to observe and measure the sonography of the tumor. The blood flow of the uterus, appendages and tumor were observed. The source of the blood flow of the tumor was measured. Doppler spectral curve was measured with multi-point sampling inside and around the lesion and the resistance index (RI), pulsatility index (PI), peak systolic velocity (PSV) and maximum temporal flow velocity (TAMXV). Multivariate Logistic regression model was used to estimate the possibility of malignancy of accessory tumor. Results: Multivariate logistic regression model can predict the degree of malignancy of tumor in accessory area more accurately than single variable (RI, tumor shape score, tumor blood flow location score, ascites) entering regression equation, the difference was significant (P < 0.000 1). Predict the probability of ovarian cancer and improve its sensitivity, specificity and positive predictive value. When the diagnostic probability is greater than 25%, the sensitivity and specificity for the diagnosis of malignant tumors, the positive predictive value and accuracy are the highest. Conclusion: Multivariate Logistic regression analysis of the diagnosis of benign and malignant tumor in the attachment zone with a high rate of compliance, multivariate Logistic regression analysis of accessory malignant tumor accuracy than the gray alone or color Doppler ultrasound alone is more diagnostic value.