探讨不同切面超声弹性成像应变比值法在诊断甲状腺结节良恶性方面的价值区别。

选取2012年11月至2013年5月因甲状腺结节疾病在解放军总医院超声科就诊的患者118例，共有153个结节，患者术前均行弹性成像检查，从纵切和横切面分别获得弹性应变率面积比值，采用ROC曲线评价其诊断价值，将结节分为临近峡部组、非临近峡部组两组，并将诊断结果与病理结果对照。

纵切面与横切面面积应变比值法在鉴别甲状腺良恶性结节的ROC曲线下面积（AUC）分别为0.906、0.844，二者无显著性差异（Z=1.54，P>0.05）；最佳诊断界点分别为：3.65、3.58；纵切面、横切面应变比值法的敏感度、特异度、准确性、阳性预测值、阴性预测值分别为：81.2%、80.8%、81.0%、87.2%、71.2%；77.2%、76.9%、77.1%、84.4%、63.4%。临近峡部组纵切面与横切面面积应变比值法在鉴别甲状腺良恶性结节的ROC曲线下面积（AUC）分别为0.903、0.830，二者无显著性差异（Z=1.125，P>0.05）；最佳诊断点分别为：3.30、3.28；纵切面、横切面应变比值法的敏感度、特异度、准确性、阳性预测值、阴性预测值分别为：86.2%、68.8%、80.0%、83.3%、73.3%；78.6%、58.8%、71.1%、75.9%、62.5%。非临近峡部组纵切面与横切面面积应变比值法在鉴别甲状腺良恶性结节的AUC分别为0.906、0.852，二者无显著性差异（Z=0.936，P>0.05）；最佳诊断界点分别为：3.33、3.71；纵切面、横切面应变比值法的敏感度、特异度、准确性、阳性预测值、阴性预测值分别为：90.1%、75.7%、85.2%、87.7%、80.0%；84.7%、70.3%、80.6%、83.6%、72.2%。

超声弹性成像有助于甲状腺结节的鉴别诊断，但横切面与纵切面的诊断效果无统计学差异。

To study the value of strain ratio from longitudinal section and cross section in differential diagnosis of benign and malignant thyroid nodules using ultrasonic elastography.

A total of 118 patients with 153 thyroid nodules underwent examinations with real-time ultrasonic elastography to obtain the strain ratio from longitudinal section and cross section. Receiver operating characteristic (ROC) curves were used to evaluate the value of elastography. The nodules are divided into two groups, group one included nodules which were near the isthmus and group two included other nodules. Diagnostic results were compared with pathological results.

The area under the ROC curve (AUC) of the strain ratios for differentiating benign and malignant thyroid nodules were 0.906 and 0.844. There was no statistical difference (Z=1.542, P>0.05). The best diagnostic cut-off points of the strain ratios were 3.65 and 3.58. The sensitivity, specificity, accuracy, positive predictive value, negative predictive value of the strain ratio from longitudinal section for differentiating benign and malignant thyroid nodules were 81.2%, 80.8%, 81.0%, 87.2% and 71.2% and those of strain ratio from cross section were 77.2%, 76.9%, 77.1%, 84.4% and 63.4%. In group one, the AUC were 0.903 and 0.830. There was no statistical difference (Z=1.125, P>0.05). The best diagnostic cut-off points were 3.30 and 3.28. The sensitivity, specificity, accuracy, positive predictive value, negative predictive value of the strain ratio from longitudinal section for differentiating benign and malignant thyroid nodules were 86.2%, 68.8%, 80.0%, 83.3% and 73.3% and those of strain ratio from cross section were 78.6%, 58.8%, 71.1%, 75.9% and 62.5%. In group two, the AUC were 0.906 and 0.852. There was no statistical difference (Z=0.936, P>0.05). The best diagnostic cut-off points were 3.33 and 3.71. The sensitivity, specificity, accuracy, positive predictive value, negative predictive value of the strain ratio from longitudinal section for differentiating benign and malignant thyroid nodules were 90.1%, 75.7%, 85.2%, 87.7% and 80.0% and those of strain ratio from cross section were 84.7%, 70.3%, 80.6%, 83.6% and 72.2%.

Real-time ultrasonic elastography is helpful to the differential diagnosis of thyroid nodules, but there is no statistical difference between the strain ratio from longitudinal section and cross section.