探讨超声支气管镜下弹性成像技术单独及联合胸部CT、常规超声在气管周围的纵隔及肺门淋巴结性质判定中的价值。

选取2016年6月至2018年3月期间在南京医科大学附属南京医院呼吸内科气管镜室拟行EBUS-TBNA检查的患者。记录目标淋巴结的胸部CT、常规超声及超声弹性成像各参数，根据EBUS-TBNA的病理阳性结果或者阴性结果进行一年的随访作为诊断的金标准。构建各个特征参数的受试者工作特征曲线，得到曲线下面积和最佳截断值，并计算出最佳截断值时各个参数的诊断准确率、灵敏度、特异度、阳性预测值、阴性预测值等指标。

共入组78例患者，117枚淋巴结：①胸部CT在判定淋巴结性质中，淋巴结短径≥10 mm，边界清楚，质地不均匀这三个特征具有统计学意义(P<0.05)，其中质地不均匀的AUC最高，为0.711；②常规超声判定淋巴结性质时，淋巴结短径≥10 mm，边界清楚，低回声，形状呈类圆形这四个特征具有统计学意义(P<0.05)，其中边界清楚的AUC最高，为0.655；③超声弹性成像图像应用图像类型、超声弹性评分、应变率比值以及蓝色面积比四种方法判断淋巴结性质，AUC分别是0.843，0.820，0.717，0.877；其中蓝色面积比的AUC最高；④联合胸部CT的淋巴结质地不均匀，常规超声的淋巴结边界清楚以及超声弹性成像的图像类型进行统计分析，得出准确率、灵敏度、特异度、阳性预测值、阴性预测值分别是85.5%，85.7%，87.9%，100%，100%，AUC为0.932。

超声弹性成像对于肺门及纵隔淋巴结的性质判断的诊断价值高于胸部CT及常规超声。三者联合能够明显提高淋巴结的恶性检出率。

To evaluate the effect of endobronchial ultrasound elastography in judging the nature of mediastinal and hilar lymph nodes by individually and jointly comparing the parameters of chest CT, ultrasound and elastography.

The persons who were admitted to the Department of Respiratory and Critical Care Medicine in Nanjing First People′s Hospital, affiliated to Nanjing Medical University for EBUS-TBNA examination from June 2016 and March 2018 were taken as the subjects for this study. An Olympus 290 bronchoscope system from Japan and an EU-ME2 ultrasound system were used. The parameters of target lymph nodes in chest CT, ultrasound and elastography were compared before EBUS-TBNA examination. The pathological results after EBUS-TBNA examination or three-month follow-up of unknown results were taken as the gold standard of diagnosis. The ROC curve of each characteristic parameter was constructed to obtain the AUC and cutoff values, and the diagnostic accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of each parameter were calculated to analyze and compare the elastography for the diagnosis of lymph node properties. In addition, the parameters of elastography for different pathological types were analyzed and statistically evaluated for further diagnostic value.

Totally 78 patients, including 117 lymph nodes, were collected. According to the differentiation of the lymph node properties by chest CT, the three properties of the lymph nodes (the short diameter≥10 mm, clear boundary and uneven texture) had statistical significant differences (P<0.05). Among them, the uneven texture of the AUC was highest, 0.711. When the characteristics of the lymph nodes were determined by ultrasound, the short diameter of the lymph nodes≥10 mm, the clear boundary, low echo and the circular shape had statistical significant difference (P<0.05). Among them, the clear boundary of the AUC was highest, 0.655. Ultrasound elastography images used image types, ultrasound elasticity score, strain rate ratio and blue area ratio to determine the nature of lymph nodes, and the values of AUC were 0.843, 0.820, 0.717, and 0.877, respectively. Among them, the blue area ratio was the highest. Inhomogeneous lymph node texture in chest CT, clear boundary of lymph nodes in ultrasound combined with elastography image types were statistically analyzed, and the accuracy, sensitivity, specificity, positive predictive value, and negative predictive value were 85.5%. 85.7%, 87.9%, 100%, 100%, respectively. And the value of AUC was 0.932. The diagnostic value was significantly higher than any single factor.

The diagnostic value of ultrasound elastography for differentiating the nature of hilar and mediastinal lymph nodes is higher than that of chest CT or ultrasound. The combination of the three techniques can significantly increase the malignant detection rate of lymph nodes.