【摘要】 目的建立应用高频彩色多普勒超声对乳腺肿块良恶性鉴别诊断的Logistic回归模型。方法采用病例对照研究方法回顾性分析经病理证实的216例乳腺肿块患者,其中良性144例,恶性72例。比较两组彩色多普勒声像图特征,对其超声特征进行多因素回归分析,建立二分类Logistic回归模型。结果逐步似然比法:乳腺良性肿块边界比恶性组清晰(P<0.05),恶性肿块无明显包膜,有蟹足或毛刺,检出率较良性组高(P<0.01),后方回声改变为回声增强或无改变和衰减,两组检出率差异有统计学意义(P<0.05)。两组之间肿块形态,内部回声,回声是否均匀,砂粒微钙化,腋窝是否有淋巴结,血流差异无统计学意义(P>0.05)。多因素回归分析显示最后进入Logistic模型的5个特征分别为形态、边界、包膜、蟹足或毛刺及后方回声改变。结论应用高频彩色多普勒超声对乳腺肿块良恶性鉴别的Logistic回归模型有助于鉴别乳腺良恶性肿块。更多还原

【Abstract】 Objective To establish a Logistic regression model for differential diagnosis of benign and malignant breast masses by high frequency and color doppler ultrasound. Methods Two hundred and sixteen cases of breast masses diagnosed pathologically were analyzed retrospectively, one hundred and forty-four cases were benign, and the remaining seventy-two cases were malignant. The ultrasonographic parameters of benign and malignant cases were recorded and compared using multiple factors binary Logistic regression analysis. Results Benign masses had a more clearer dividing line than malignant masses(P<0.05). Malignant masses manifested more clearly crab foot or burr, without envelope compared with benign masses(P<0.01), rear echo change was echo enhancement or no change and attenuation, two groups of detection rate had a significant difference(P<0.05). There was no statistically significant difference between benign and malignant masses in shape, internal echo, echo was homogeneous, sand point like calcification, whether there were lymph nodes in axillary, blood flow(P>0.05). Five ultrasonic features were finally implemented into the Logistic regression model which included boundary, envelope, crab foot or burr and rear echo change. Conclusion The Logistic regression model can be helpful for differentiation of benign or malignant breast masses.更多还原