论文部分内容阅读
目的:探讨超声弹性对比指数（ECI）在甲状腺良恶性结节鉴别诊断中的价值，并对比不同测量方法的效能。方法:选取2018年5月至2019年5月滨州医学院附属医院甲状腺结节患者122例（131个结节）。用横切面结节内（横内）、横切面结节与周围组织交界处（横界）、纵切面结节内（纵内）和纵切面结节与周围组织交界处（纵界）四种方法测量ECI，采用受试者工作特征（ROC）曲线分析不同测量方法诊断甲状腺恶性结节的最佳ECI截断值。结果:131个结节中，良性54个，恶性77个。良性结节横内、横界、纵内和纵界测量的ECI均明显低于恶性结节[2.10（1.48，2.34）比3.07（2.73，3.87）、1.91（1.64，2.18）比2.62（2.24，3.07）、2.19（1.59，2.39）比3.00（2.72，3.63）和1.89（1.71，2.16）比2.66（2.21，3.10）]，差异有统计学意义（n P<0.05）。四种测量方法中，横内测量曲线下面积最大，为0.925，对应的截断值为2.43，灵敏感、特异度、准确率、阳性预测值和阴性预测值分别为90.9%、87.0%、88.5%、89.7%和86.8%。n 结论:ECI可辅助常规超声鉴别诊断甲状腺结节性质，其中横内测量ECI效能最佳。“,”Objective:To explore the value of the elasticity contrast index (ECI) in differential diagnosis of thyroid nodules, and contrast the effect of 4 different methods on measuring ECI.Methods:A total of 122 patients with 131 thyroid nodules in Binzhou Medical University Hospital from May 2018 to May 2019 were enrolled, and elastography was performed in 4 different ways such as in axial plane for internal (AI), axial plane for periintranodular (AP), longitudinal plane for internal (LI) and longitudinal plane for periintranodular (LP). The cut-off values for predicting malignant nodules in 4 different ways were determined separately using receiver operating characteristic (ROC) curve analysis.Results:There were 54 benign and 77 malignant ones in 131 nodules. The ECI in AI, AP, LI and LP in benign thyroid nodules was significantly lower than that in malignant ones: 2.10 (1.48, 2.34) vs. 3.07 (2.73, 3.87), 1.91 (1.64, 2.18) vs. 2.62 (2.24, 3.07), 2.19 (1.59, 2.39) vs. 3.00 (2.72, 3.63) and 1.89 (1.71, 2.16) vs. 2.66 (2.21, 3.10), and there was statistical difference (n P<0.05). Among the 4 different ways, the largest area under curve was achieved in AI with 0.925, and the corresponding optimal diagnostic threshold was 2.43 (with 90.9% sensitivity, 87.0% specificity, 88.5% accuracy, 89.7% positive predictive value and 86.8% negative predictive value).n Conclusions:ECI is helpful for conventional ultrasound to diagnose thyroid nodules as malignant or benign. The best value is obtained in AI.