【摘要】 目的:探索基于3D-SMI及SWE定性定量参数,应用二元Logistic回归分析是否可以提高乳腺肿块的诊断效能。方法:回顾性研究2020年11月—2021年2月在我院就诊的193例乳腺肿块患者200个肿块的临床资料。所有患者经过常规超声、超微血管三维成像(3D-SMI)、剪切波弹性成像(SWE)检查并进行细针穿刺抽吸细胞学检查(FNA)或手术切除,证实为乳腺癌患者77例。首先应用乳腺超声影像报告规范与数据系统(BI-RADS)分类指标和病理结果建立常规超声二元Logistic回归模型,然后联合3D-SMI、SWE定性定量参数再建回归模型。应用受试者工作特征曲线(ROC)比较两种回归模型的诊断效能。结果:基于3D-SMI及SWE定性定量参数的二元Logistic回归模型对乳腺肿块诊断的灵敏度、特异度、准确度较常规超声模型明显提高。结论:基于3D-SMI及SWE定性定量参数在诊断乳腺肿块的二元Logistic回归分析可以提高超声的诊断效能。更多还原

【Abstract】 Objective To explore whether binary logistic regression analysis based on qualitative and quantitative parameters of 3 D-SMI and SWE can improve the diagnostic efficiency of breast masses. Methods Clinical data of 193 patients with 200 breast masses admitted to our hospital from November 2020 to February 2021 were retrospectively studied. All patients were examined by conventional ultrasound imaging, three-dimensional superb microvascular imaging, shear wave elastography, and all of breast masses were confirmed by fine-needle aspiration biopsy or surgical removal, breast cancers were confirmed in 77 patients. First, the binary Logistic regression model of conventional ultrasound was established based on the classification indexes of BI-RADS and pathological results, and then another regression model was established by combining the qualitative and quantitative parameters of 3 D-SMI SWE. The diagnostic efficiency of two models were compared by receiver operating characteristic curve. Results The sensitivity, specificity and accuracy of binary Logistic regression model based on qualitative and quantitative parameters of 3 D-SMI and SWE in the diagnosis of breast masses were significantly improved compared with conventional ultrasound regression model.Conclusion The application of Logistic regression analysis based on qualitative and quantitative parameters of 3 D-SMI and SWE in the diagnosis of breast masses can improve the ultrasonic diagnosis efficiency.更多还原