声触诊组织成像与定量技术联合2015年美国甲状腺学会指南在甲状腺良恶性结节鉴别诊断中的应用研究

Study on the Application of Virtual Touch Tissue Imaging Quantification in Combination With the 2015 Guidelines of American Thyroid Association for Differentiating Benign Thyroid Nodules From Malignant Thyroid Nodules

目的探讨声触诊组织成像与定量(VTIQ)技术联合2015年美国甲状腺学会(ATA)指南在甲状腺良恶性结节鉴别诊断中的临床应用价值。方法选取2015年5—11月同济大学附属第十人民医院收治的甲状腺结节患者109例,共109个甲状腺结节。患者行常规甲状腺超声检查和VTIQ检查,并按照2015年ATA指南进行分组。以病理组织学检查结果为金标准,比较单独应用VTIQ检查、ATA指南以及两者联合应用诊断甲状腺良恶性结节的价值。结果 109个甲状腺结节中,良性结节65个,恶性结节44个。良恶性结节大小比较,差异无统计学意义(P>0.05);良亚性结节内部成分、回声、边缘、钙化、纵横比比较,差异均有统计学意义(P<0.001)。良性结节剪切波速度(SWS)最大值、最小值、平均值均低于恶性结节(P<0.001);良恶性结节2015年ATA指南分组情况比较,差异有统计学意义(P<0.001)。SWS最大值、最小值、平均值诊断甲状腺良恶性结节的受试者工作特征(ROC)曲线下面积(AUC)分别为0.834〔95%CI(0.758,0.911)〕、0.841〔95%CI(0.763,0.919)〕和0.861〔95%CI(0.790,0.933)〕,诊断界值分别为3.26、2.54、2.92 m/s。以SWS平均值2.92 m/s为诊断界值时,其诊断甲状腺良恶性结节的灵敏度、特异度及正确率分别为79.5%(35/44)、83.1%(54/65)和81.7%(89/109)。以极低度可疑恶性组诊断为甲状腺良性结节,低度可疑恶性组及以上(包括未描述组)诊断为甲状腺恶性结节,其诊断甲状腺良恶性结节的灵敏度、特异度及正确率分别为97.7%(43/44)、40.0%(26/65)和63.3%(69/109),AUC为0.689〔95%CI(0.591,0.786)〕。将极低度可疑恶性组诊断为甲状腺良性结节,高度可疑恶性组诊断为甲状腺恶性结节,低度、中度可疑恶性组及未描述组则根据SWS平均值<2.92 m/s诊断为甲状腺良性结节,SWS平均值≥2.92 m/s诊断为甲状腺恶性结节。VTIQ技术联合2015年ATA指南诊断甲状腺良恶性结节的灵敏度、特异度及正确率分别为97.7%(43/44)、80.0%(52/65)、87.2%(95/109),AUC为0.889〔95%CI(0.824,0.954)〕。VTIQ技术联合2015年ATA指南诊断甲状腺良恶性结节的灵敏度与单独应用2015年ATA指南相同,而特异度和正确率均升高(P<0.001)。VTIQ技术联合2015年ATA指南诊断甲状腺良恶性结节的灵敏度较单独应用VTIQ技术升高(P=0.008),而特异度和正确率比较,差异均无统计学意义(P=0.727,P=0.210)。结论 VTIQ技术对于鉴别甲状腺结节的良恶性具有重要的临床价值,联合2015年ATA指南后较单独应用2015年ATA指南能明显提高诊断的特异度和正确率,同时不降低灵敏度。

Objective To assess the clinical application value of virtual touch tissue imaging quantification（ VTIQ）in combination with the 2015 guidelines of American Thyroid Association（ ATA） for differentiating benign thyroid nodules from malignant thyroid nodules. Methods 109 patients with thyroid nodules in the Tenth People＇ s Hospital Affiliated to Tongji University from May to November in 2015,were selected as study subjects,109 thyroid nodules were confirmed by pathology after surgery. Patients underwent conventional ultrasonic examinations and VTIQ examinations,and were grouped according to the2015 guidelines of ATA. Histopathological examination results were the gold standard,the values of simple application of VTIQ,simple application of 2015 guidelines of ATA, and application of VTIQ in combination with 2015 guidelines of ATA for differentiating malignant thyroid nodules from benign thyroid nodules were compared. Results There were 65 benign thyroid nodules and 44 malignant thyroid nodules among 109 thyroid nodules. There was not significant difference in diameter between benign thyroid nodules and malignant thyroid nodules（ P〈0. 05）; there were significant differences in ingredient, echo,nodule edge,calcification and aspect ratio between benign thyroid nodules and malignant thyroid nodules（ P〈0. 001）. The maximum value of SWS,the minimum value of SWS,and the average value of SWS of benign thyroid nodules were all lower than those of malignant thyroid nodules（ P〈0. 001）. There was significant difference in 2015 guidelines of ATA grouping situation between benign thyroid nodules and malignant thyroid nodules（ P〈0. 001）. For differentiating benign thyroid nodules from malignant thyroid nodules,the areas under the ROC curves（ AUC） for the maximum value of SWS,the minimum value of SWS and the average value of SWS was 0. 834 〔95% CI（ 0. 758,0. 911） 〕,0. 841 〔95% CI（ 0. 763,0. 919） 〕and 0. 861 〔95%CI（ 0. 790,0. 933） 〕,respectively,the diagnostic cut- off value was 3. 26,2. 54 and 2. 92 m / s,respectively. With a cut-off value of 2. 92 m / s,the sensitivity,specificity and accuracy of the average value of SWS were 79. 5%（ 35 /44）,83. 1%（ 54 /65） and 81. 7%（ 89 /109）,respectively. Cases in very low suspicious malignant group were diagnosed with benign thyroid nodules,cases in low suspicious malignant group and above（ including no description group） were diagnosed with malignant thyroid nodules,the sensitivity,specificity and accuracy was 97. 7%（ 43 /44）,40. 0%（ 26 /65） and 63. 3%（ 69 /109）respectively in differentiating malignant thyroid nodules from benign thyroid nodules,the AUC was 0. 689 〔95% CI（ 0. 591,0. 786） 〕. Cases in very low suspicious malignant group were diagnosed with benign thyroid nodules,cases in highly suspicious for malignancy group were diagnosed with malignant thyroid nodules; for low- grade,moderate suspicious malignant group and no description group,the diagnosis of benign thyroid nodules was made according to the SWS average value 2. 92 m / s,and the diagnosis of malignant thyroid nodules was made according to the SWS average value≥2. 92 m / s. The sensitivity,specificity and accuracy of VTIQ in combination with the 2015 guidelines of ATA in the diagnosis of benign thyroid nodules and malignant thyroid nodules was 97. 7%（ 43 /44）,80. 0%（ 52 /65）,87. 2%（ 95 /109）,respectively,the AUC was 0. 889 〔95% CI（ 0. 824,0. 954） 〕. The sensitivity of VTIQ in combination with the 2015 guidelines of ATA was equal to simple application of 2015 guidelines of ATA,while the specificity and accuracy of VTIQ in combination with the 2015 guidelines of ATA were significantly higher than those of simple application of 2015 guidelines of ATA, respectively（ P〈0. 001）. The sensitivity of VTIQ in combination with the 2015 guidelines of ATA was significantly higher than that of simple application of VTIQ（ P = 0. 008）,there was no significant difference in the specificity and accuracy between VTIQ in combination with the 2015 guidelines of ATA and VTIQ（ P = 0. 727,P = 0. 210）. Conclusion VTIQ is an important method for differentiating malignant thyroid nodules from benign thyroid nodules. Compared with simple application of 2015 guidelines of ATA,application of VTIQ in combination with2015 guidelines of ATA can improve the specificity and accuracy of thyroid nodules diagnosis without loss of sensitivity.