声触诊组织成像定量技术诊断三阴性乳腺癌

Virtual Touch Tissue Imaging Quantification Diagnose Triplenegative Breast Cancer

目的探究声触诊组织成像定量技术(virtual touch tissue imaging quantification,VTIQ)在三阴性乳腺癌(triple negative breast cancer,TNBC)中的诊断价值.方法选择2017年1月~2018年6月于笔者医院就诊的乳腺肿块患者217例(共255个肿块),所有患者术前均分别进行乳腺常规二维超声、超声弹性成像(elasticity imaging,EI)及VITQ检查,以病理诊断作为金标准,比较不同病理性质乳腺肿块患者3种超声检测方法指标的差异.绘制受试者工作曲线(receiver operating characteristic curve,ROC)并评价3种超声检测方法指标判断三阴性乳腺癌的诊断效能(包括诊断准确率、敏感度、特异性、阳性预测值和阴性预测值等).结果根据病理结果,共检出乳腺良性病变140个,非TNBC 90个,TNBC 25个.常规超声结果提示:TNBC的BI-RADS量化分值显著高于乳腺良性病变和非TNBC(F=-11.241,P=0.000);EI结果提示,TNBC的改良5分法分值显著高于乳腺良性病变和非TNBC(F=-10.627,P=0.000);VTIQ结果提示,TNBC的SWV最大值、最小值及平均值显著高于乳腺良性病变和非TNBC(F=-18.439、-12.080、-9.873,P=0.000).ROC曲线显示,SWV最大值、SWV最小值和SWV平均值在预测TNBC中的AUC(AUC =0.922、0.938、0.931)显著高于EI(AUC =0.906)和常规超声(AUC=0.815),其中,SWV最大值、SWV最小值和SWV平均值预测TNBC的最佳截点分别为≥5.131m/s、≥3.562m/s和≥4.279m/s.应用ROC曲线所得出的最佳截点,SWV最大值、SWV最小值和SWV平均值在预测三阴性乳腺癌时的诊断准确率(93.28%、92.63%、93.58%)、敏感度(92.13%、91.17%、93.33%)、特异性(91.35%、93.44%、91.27%)和阳性预测值(92.65%、91.61%、90.92%)、阴性预测值(96.16%、93.15%、94.12%)均明显高于EI和常规超声.结论 VTIQ在预测TNBC中具有较高价值,其中SWV最大值≥5.131m/s、SWV最小值≥3.562m/s和SWV平均值≥4.279m/s可作为预测TNBC的参考指标.

Objective To explore the diagnostic value of virtual touch tissue imaging quantification in the triplenegative breast cancer. Methods Totally 217 patients(255 breast lumps) who suffered breast tumors from January 2017 to June 2018 in our hospital were chosen and checked by conventional two - dimensional ultrasound, elasticity imaging( EI) and VITQ before surgery. According to the gold standard of pathology, the different indexes of these three kinds of ultrasound techniques among patients with different kinds of breast lumps pathological results were compared. The receiver operating characteristic curves ( ROC ) were plotted and the diagnostic efficiency (including diagnostic accuracy, sensitivity, specificity, positive predictive value and negative predictive value, etc.) of three kinds of ultrasound techniques on TNBC diagnosis were evaluated. Results According to pathological results, there were 140 cases of breast benign lesions , 90 cases of non 一 TNBC breast cancers and 25 cases of TNBCs. Conventional ultrasound showed that the quantified scores of BI - RADS in TNBCs were significantly higher than those in breast benign lesions and non - TNBC breast cancers ( F =-11. 241;P = 0. 000). EI showed that the scores based on the improved 5 - score system in TNBCs were significantly higher than those in breast benign lesions and non - TNBC breast cancers( F =- 10. 627 , P = 0. 000 ). VITQ showed that the SWV maximum value, SWV minimum value and SWV average value in TNBCs were significantly higher than those in breast benign lesions and non - TNBC breast cancers ( F =-18.439 ,- 12.080,-9. 873 ,P = 0. 000 ). The ROC curve showed that the AUC of SWV maximum value, SWV minimum value and SWV average value in predicting TNBCs( AUC = 0. 922 ,0. 938 ,0. 931 ) were significantly higher than EI( AUC = 0. 906) and conventional ultrasound ( AUC =0.815), of which the best cut - off point for SWV maximum value, SWV minimum value and SWV average value were M5. 131m/s、M3. 562m/s and M4. 279m/s. The diagnostic accuracy(93. 28%, 92. 63%, 93. 58%),sensitivity(92. 13%, 91. 17%,93.33%), specificity (91.35%, 93. 44%,91.27%) and positive predictive value ( 92. 65%, 91.61 %, 90. 92%), negative predictive value( 96. 16%,93. 15%,94. 12%) of SWV maximum value, SWV minimum value and SWV average value in predicting TNBC were significantly higher than those of El and conventional ultrasound. Conclusion VITQ is valuable in the prediction of TNBC. SWV maximum value M5. 131 m/s,SWV minimum value562m/s and SWV average value^4. 279m/s can be used as reference indexes for the prediction of TNBC.