Zhang Fengjuan, Zhao Xinming, Han Ruoling, Du Min, Li Ping, Ji Xiaohui
Departments of Medical Ultrasound, the Fourth Hospital of Hebei Medical University, Hebei, Shijiazhuang, China.
Department of Nuclear Medicine and Medical Imaging, the Fourth Hospital of Hebei Medical University, Hebei, Shijiazhuang, China.
J Ultrasound Med. 2017 Dec;36(12):2533-2543. doi: 10.1002/jum.14302. Epub 2017 Jun 24.
This study aimed to evaluate the diagnostic value of acoustic radiation force impulse (ARFI) imaging and strain elastography in differentiating malignant from benign thyroid nodules.
This study included 128 patients (104 female and 24 male; mean age ± standard deviation, 48 ± 11 years; range, 23-76 years) with 152 pathologically proven thyroid nodules. Conventional sonography, strain ratio (SR) of strain elastography, virtual touch tissue imaging (VTI) grade, and virtual touch tissue quantification (shear wave velocity [SWV]) from ARFI imaging data were generated for each nodule. The diagnostic performance of the three elastography analysis methods was analyzed and compared by multiple receiver operating characteristic curve analysis. The interobserver agreement for VTI grade was also assessed.
The best cutoff values for VTI grade, mean SWV, and mean SR for predicting malignant thyroid nodules were greater than or equal to grade χ, 2.87 m/s and 2.37, respectively. The area under the receiver operating characteristic curve for VTI grade, mean SWV, and mean SR was 0.854, 0.869 and 0.849, respectively (P > .05), and the accuracy was 83.55, 81.58 and 80.26%, respectively (P > .05). The accuracy of the combined use of conventional sonography and ARFI imaging for VTI grade, mean SWV, and mean SR was 98.03, 95.39 and 96.71%, respectively, which was slightly higher than that of conventional sonography (P > .05). The interobserver agreement produced a κ-value of 0.98 (95% confidence interval, 0.959-1.000).
Both strain elastography and ARFI imaging have high sensitivity and specificity for differentiating malignant from benign thyroid nodules. Combined with sonography, these techniques can improve the accuracy of thyroid nodule diagnosis.
本研究旨在评估声辐射力脉冲（ARFI）成像和应变弹性成像在鉴别甲状腺良恶性结节中的诊断价值。
本研究纳入了128例患者（104例女性，24例男性；平均年龄±标准差，48±11岁；范围，23 - 76岁），共152个经病理证实的甲状腺结节。对每个结节进行常规超声检查、应变弹性成像的应变率（SR）、虚拟触诊组织成像（VTI）分级以及基于ARFI成像数据的虚拟触诊组织定量（剪切波速度[SWV]）。通过多个受试者操作特征曲线分析来分析和比较三种弹性成像分析方法的诊断性能。还评估了观察者间对VTI分级的一致性。
预测甲状腺恶性结节的VTI分级、平均SWV和平均SR的最佳截断值分别为大于或等于χ级、2.87 m/s和2.37。VTI分级、平均SWV和平均SR的受试者操作特征曲线下面积分别为0.854、0.869和0.849（P＞0.05），准确率分别为83.55%、81.58%和80.26%（P＞0.05）。常规超声与ARFI成像联合用于VTI分级、平均SWV和平均SR的准确率分别为98.03%、95.39%和96.71%，略高于常规超声（P＞0.05）。观察者间一致性的κ值为0.98（95%置信区间，0.959 - 1.000）。
应变弹性成像和ARFI成像在鉴别甲状腺良恶性结节方面均具有较高的敏感性和特异性。与超声检查相结合，这些技术可提高甲状腺结节诊断的准确性。
