Liu Bo-Ji, Lu Feng, Xu Hui-Xiong, Guo Le-Hang, Li Dan-Dan, Bo Xiao-Wan, Li Xiao-Long, Zhang Yi-Feng, Xu Jun-Mei, Xu Xiao-Hong, Qu Shen
Department of Medical Ultrasound, Shanghai Tenth People's Hospital, Ultrasound Research and Education Institute, Tongji University School of Medicine Shanghai 200072, China ; Thyroid Institute, Tongji University School of Medicine Shanghai 200072, China.
Department of Medical Ultrasound, Shanghai Tenth People's Hospital, Ultrasound Research and Education Institute, Tongji University School of Medicine Shanghai 200072, China ; Thyroid Institute, Tongji University School of Medicine Shanghai 200072, China ; Department of Ultrasound, Guangdong Medical College Affiliated Hospital Zhanjiang 524001, China.
Int J Clin Exp Med. 2015 Sep 15;8(9):15362-72. eCollection 2015.
The aim of this study was to evaluate the potential diagnostic performance of acoustic radiation force impulse (ARFI) elastography in identifying malignancy in nodules that do not appear highly suspicious on conventional ultrasound (US).
330 pathologically confirmed thyroid nodules (40 malignant and 290 benign; mean size, 22.0±11.6 mm) not suspicious of malignancy on conventional US in 330 patients (mean age 52.8±11.7 years) underwent ARFI elastography before surgery. ARFI elastography included qualitative ARFI-induced strain elastography (SE) and quantitative point shear wave elastography (p-SWE). ARFI-induced SE image was assessed by SE score, while p-SWE was denoted with shear wave velocity (SWV, m/s). The diagnostic performance of four criteria sets was evaluated: criteria set 1 (ARFI-induced SE), criteria set 2 (p-SWE), criteria set 3 (either set 1 or 2), criteria set 4 (both set 1 and 2). Receiver operating characteristic curve (ROC) analyses were performed to assess the diagnostic performance.
SE score ≥4 was more frequently found in malignant nodules (32/40) than in benign nodules (30/290, P<0.001). The mean SWV of malignant nodules (3.64±2.23 m/s) was significantly higher than that of benign nodules (2.02±0.69 m/s) (P<0.001). ARFI-induced SE (set 1) had a sensitivity of 80.0% (32/40) and a specificity of 89.7% (260/290) with a cut-off point of SE score ≥4; p-SWE (set 2) had a sensitivity of 80.0% (32/40) and a specificity of 57.9% (168/290) with a cut-off point of SWV ≥2.15 m/s. When ARFI-induced SE and p-SWE were combined, set 3 had the highest sensitivity (92.5%, 37/40) while set 4 had the highest specificity (95.2%, 276/290).
ARFI elastography can be used for differential diagnosis of malignant thyroid nodules without highly suspicious features on US. The combination of ARFI-induced SE and p-SWE leads to improved sensitivity and specificity.
本研究旨在评估声辐射力脉冲（ARFI）弹性成像技术在鉴别常规超声（US）表现无高度可疑特征的结节良恶性方面的潜在诊断性能。
330例患者（平均年龄52.8±11.7岁）的330个经病理证实的甲状腺结节（40个恶性，290个良性；平均大小22.0±11.6mm），这些结节在常规US检查中无恶性可疑表现，术前均接受了ARFI弹性成像检查。ARFI弹性成像包括定性的ARFI诱导应变弹性成像（SE）和定量的点剪切波弹性成像（p-SWE）。ARFI诱导的SE图像通过SE评分进行评估，而p-SWE则用剪切波速度（SWV，m/s）表示。评估了四组标准的诊断性能：标准组1（ARFI诱导的SE）、标准组2（p-SWE）、标准组3（标准组1或标准组2）、标准组4（标准组1和标准组2）。采用受试者操作特征曲线（ROC）分析来评估诊断性能。
SE评分≥4在恶性结节（32/40）中比在良性结节（30/290）中更常见（P<0.001）。恶性结节的平均SWV（3.64±2.23m/s）显著高于良性结节（2.02±0.69m/s）（P<0.001）。ARFI诱导的SE（标准组1）以SE评分≥4为截断点时，灵敏度为80.0%（32/40），特异度为89.7%（260/290）；p-SWE（标准组2）以SWV≥2.15m/s为截断点时，灵敏度为80.0%（32/40），特异度为57.9%（168/290）。当将ARFI诱导的SE和p-SWE联合使用时，标准组3的灵敏度最高（92.5%，37/40），而标准组4的特异度最高（95.2%，276/290）。
ARFI弹性成像可用于鉴别US表现无高度可疑特征的甲状腺恶性结节。ARFI诱导的SE和p-SWE联合使用可提高灵敏度和特异度。
