论文部分内容阅读
目的建立细胞形态计量学对胸腹水鉴别诊断的Logistic回归模型。方法回顾性分析60例患者胸腹水的细胞形态参数,包括胞体大小、胞质特征、胞核大小、核质比、核仁/核、细胞核分裂象、细胞分布特点,通过多因素回归分析建立二分类Logistic回归模型,评价回归模型预报良性、恶性胸腹水的效能。结果进入Logistic回归模型的5个参数分别为胞体大小、胞核大小、核质比、核仁/核、细胞分布特点,回归模型预报良性、恶性胸腹水的准确度、敏感度、特异度分别为90.0%、86.7%、93.3%,ROC曲线下面积为0.977。结论胸腹水细胞形态计量学的Logistic回归模型有助于胸腹水的良性、恶性鉴别诊断。
Objective To establish a Logistic regression model for differential diagnosis of pleural effusion and ascites by cell morphology. Methods The morphological parameters of pleural effusion and ascites fluid in 60 patients were retrospectively analyzed, including cell body size, cytoplasmic characteristics, nucleus size, nuclear to cytoplasm ratio, nucleolus / nucleus, mitosis, and cell distribution. The multivariate regression analysis was used to establish two Classification Logistic regression model to evaluate the regression model to predict the efficacy of benign and malignant pleural effusion. Results Logistic regression model of the five parameters were cell size, nucleus size, nuclear mass ratio, nucleolus / nucleus, cell distribution characteristics, regression model prediction of benign and malignant pleural effusion accuracy, sensitivity and specificity were 90.0%, 86.7%, 93.3%, and the area under the ROC curve was 0.977. Conclusions Logistic regression model of pleural and ascitic fluid cytomorphometry is helpful for the differential diagnosis of benign and malignant pleural effusion.