声触诊组织成像定量技术在乳腺癌早期筛查中的应用价值

Application of Virtual Touch Tissue Imaging Quantification in Early Screening of Breast Cancer

目的探讨声触诊组织成像定量(Virtual Touch Tissue Imaging Quantification,VTIQ)技术在乳腺癌早期筛查以及乳腺肿块良恶性鉴别诊断方面的应用价值。方法回顾性分析经健康体检发现并经术后病理证实的162例乳腺肿瘤患者(均单发)的常规超声、VTIQ技术及临床资料,随机分为训练组(114例)和验证组(48例)。VTIQ检查获得肿块剪切波速度(Shear Wave Velocity,SWV)最大值、最小值和平均值。以病理结果作为金标准,绘制VTIQ技术诊断乳腺肿块良、恶性的受试者工作特征(Receiver Operator Characteristic,ROC)曲线,分析其诊断效能。训练组内各指标行多因素logistic回归分析,确定鉴别乳腺肿瘤良、恶性的独立预测指标,建立列线图模型,验证其预测准确性。结果术后病理显示162个乳腺肿块包括恶性47个、良性115个,恶性肿块(SWV)最大值、最小值和平均值均高于良性肿块。多因素logistic回归分析显示,SWV最大值(OR=21.086,95%CI:2.222~200.122,P=0.008)和患者年龄(OR=1.126,95%CI:1.054~1.203,P<0.01)是鉴别乳腺肿瘤良、恶性的独立预测指标,据此建立的列线图预测模型具有良好的预测准确性(训练组AUC为0.964,验证组为0.973)。预测模型AUC(0.964)显著高于SWV最大值AUC(P<0.05)和患者年龄AUC(0.814,P<0.01)。结论VTIQ技术能够很好地辅助常规超声进行乳腺癌早期筛查以及良、恶性鉴别。经VTIQ获得的SWV最大值联合患者年龄建立的列线图模型具有良好的预测准确性,具有一定的临床应用价值。

Objective To investigate the diagnostic value of virtual touch tissue imaging quantification(VTIQ)in differential diagnosis of benign and malignant breast masses.Methods The data of 162 breast patients(all single)found by physical examination and confirmed by postoperative pathology were analyzed retrospectively and randomly divided into training group(114 cases)and verification group(48 cases).Based on the pathological results of surgery as the gold standard,the receiver operator characteristic(ROC)curve of subjects diagnosed with VTIQ for benign and malignant breast masses was drawn to analyze its diagnostic efficiency.The maximum,minimum and mean values of shear wave velocity(SWV)were obtained by VTIQ.Multivariate logistic regression analysis was performed to determine the independent predictor in the training cohort for building a nomogram,and the nomogram’s predictive accuracy was validated using the area under the receiver operating characteristic curve(AUC).Results Among the 162 breast masses,115 benign lesions and 47 carcinomas were diagnosed by pathology.The maximum,minimum and mean values of SWV for carcinomas were higher than those for benign masses.Postoperative pathology showed 162 breast masses,including 47 malignant and 115 benign.The maximum,minimum and median SWV of malignant masses were higher than those of benign masses.Multivariate logistic regression analysis showed that the maximum value of SWV(OR=21.086,95%CI:2.222-200.122,P=0.008)and patient age(OR=1.126,95%CI:1.054-1.203,P<0.01)were independent predictors to distinguish benign and malignant breast tumors.The nomogram prediction model established based on this had good prediction accuracy(AUC of training group was 0.964 and validation group was 0.973).The AUC of prediction model(0.964)was significantly higher than the maximum AUC of SWV(P<0.05)and the AUC of patient age(0.814,P<0.01).Conclusion VTIQ can well assist conventional ultrasound in early screening and differentiating benign and malignant breast masses.Adding patient age to the maximum value of SWV could improve its diagnostic sensitivity for breast masses.The newly-constructed nomogram based on the maximum value of SWV obtained from VTIQ technique and patient age exhibits satisfactory predictive accuracy.