超声造影在甲状腺结节鉴别诊断中的价值

目的探讨超声造影对甲状腺影像报告和数据系统(TI-RADS)分类的甲状腺结节进一步鉴别诊断的价值。 方法对295例387个常规超声TI-RADS分类的甲状腺结节行超声造影检查，观察造影后甲状腺结节的微循环灌注特征和应用SonoLiver CAP软件获得的时间-强度曲线各定量参数分析，综合进行两次TI-RADS分类评估，并与术后病理结果进行对照，计算常规超声TI-RADS、超声造影后TI-RADS分类诊断甲状腺恶性结节的敏感度、特异度、准确度、阳性预测值、阴性预测值，并绘制常规超声TI-RADS、超声造影后二次TI-RADS分类诊断甲状腺恶性结节的受试者操作特性(ROC)曲线，计算ROC曲线下面积。 结果295例387个常规超声TI-RADS分类2～5类的甲状腺结节，术后病理结果为恶性82例121个、良性213例266个。常规超声TI-RADS分类诊断恶性结节的敏感度63.64 %(77/121)、特异度73.87%(229/310)、准确度79.07%(306/387)、阳性预测值45.83%(77/168)、阴性预测值83.88%(229/273)，而超声造影后二次TI-RADS分类诊断恶性结节的敏感度85.95%(104/121)、特异度92.93%(263/283)、准确度94.83%(263/387)、阳性预测值83.87%(104/124)、阴性预测值93.93%(263/280)，超声造影后二次TI-RADS分类诊断甲状腺恶性结节的敏感度、特异度、准确度、阳性预测值、阴性预测值均高于常规超声TI-RADS分类诊断，差异有统计学意义(χ² ＝5.92，4.38，7.31，7.45，3.13；P<0.05)。常规超声TI-RADS、超声造影后二次TI-RADS分类诊断甲状腺恶性结节的ROC曲线下面积分别为0.753，0.834。二次TI-RADS分类评估后甲状腺结节恶性组的超声造影增强模式：≥1 cm实性结节早期快速不均匀增强伴灌注缺损区，<1 cm实性结节呈不均匀低增强；实性结节良性组增强模式表现为均匀高或等增强以及结节周边环状高增强，两者比较差异有统计学意义(χ² ＝43.86，12.54；P<0.05)；良恶性组间在增强水平、增强边界、始退特点、相对平均渡越时间、峰值强度、平均强度均具有统计学意义(χ² ＝51.48，17.91，47.25，8.92，3.37，4.05；P<0.05)；而增强范围、平均强度、峰值强度在良恶性组间均无统计学意义(χ² ＝2.87，0.95，0.88；P>0.05)。 结论超声造影对常规超声TI-RADS分类的甲状腺结节可以做进一步良恶性鉴别，提高术前诊断准确率。

Differential diagnostic value of contrast-enhanced ultrasonography in thyroid nodules

ObjectiveTo investigate the value of contrast-enhanced ultrasonography(CEUS) in the diagnosis of thyroid nodules under Thyroid Imaging Reporting and Data System (TI-RADS) classification. MethodsTotally 295 patients were enrolled, 387 thyroid nodules underwent conventional ultrasonic TI-RADS classification CEUS.In order to observe the microcirculation perfusion characteristics of thyroid nodules, the time intensity curve was obtained by CAP SonoLiver software and compare with postoperative pathologic results.The quantitative parameters of the intensity curves were analyzed, and the two TI-RADS classification were performed.The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of diagnosis of thyroid malignant nodules were calculated by conventional ultrasonic TI-RADS as well as the two TI-RADS classification.Meanwhile the TI-RADS (ROC) curves were plotted, and the area under the ROC curve was calculated. ResultsThree hundred and eight-seven conventional classification of TI-RADS 2~5 classes of thyroid nodules in 295 patients′ postoperative pathology were analyzed, 121 cases in 82 patients were malignant and 213 cases in 266 patients were benign.The sensitivity, specificity, accuracy, positive predictive value and negative predictive value of conventional ultrasound TI-RADS classification for diagnosis of malignant nodules were 63.64% (77/121), 73.87% (229/310), 79.07% (306/387), 45.83% (77/168), and 83.88% (229/273), while the two TI-RADS classification for diagnosis of malignant nodules were 85.95% (104/121), 92.93% (263/283), 94.83% (263/387), 83.87% (104/124), and 93.93% (263/280). The sensitivity, specificity, accuracy, positive predictive value, positive predictive value and negative predictive value of the two TI-RADS classification of thyroid malignant nodules were higher than those of conventional ultrasound(χ²=5.92, 4.38, 7.31, 7.45, 3.13; P<0.05). The area under the ROC curve of conventional ultrasonic and the two TI-RADS classification of thyroid malignant nodules were 0.753 and 0.834, respectively.Among cases that TI-RADS was classified as malignant thyroid nodules by CEUS, they exhibited the following features under ultrasound contrast-enhanced mode: solid nodules (≥1 cm) exhibited early rapid heterogeneous enhancement pattern together with perfusion defects, and solid nodules (<1 cm) also exhibited heterogeneous enhancement pattern.For cases that TI-RADS was classified as benign thyroid nodules by CEUS, nodules and their annular surroundings exhibited homogenous enhancement pattern.The differences in the features between malignant group and benign group were statistically significant (χ²=43.86, 12.54; P<0.05). The differences between malignant group and benign group were statistically significant (χ²=51.48, 17.91, 47.25, 8.92, 3.37, 4.05; P<0.05) in enhancement level, enhancement border, arrival to peak-time characteristics, quantitative parameters mean transit time(MTT), peak intensing(PI) and arrival intensing(AI), (χ²=51.48, 17.91, 47.25, 8.92, 3.37, 4.05; P<0.05), but not statistically significant in enhancement value range, quantitative parameters rise time(RT) and peak time(PT) (χ²=2.87, 0.95, 0.88; P>0.05). ConclusionCEUS can improve accuracy in TI-RADS classification of thyroid nodules.