人工智能辅助评分联合剪切波弹性成像鉴别诊断甲状腺良恶性结节的临床价值

Clinical value of artificial intelligence assisted score combined with shear wave elastography in the differential diagnosis of benign and malignant thyroid nodules

目的探讨人工智能辅助评分(AIAS)联合剪切波弹性成像(SWE)鉴别诊断甲状腺良恶性结节的临床价值。方法选取我院经手术或穿刺活检病理确诊的甲状腺结节患者231例,共检出327个结节,其中良性结节135个,恶性结节192个,比较甲状腺良恶性结节SWE参数[应变率比值(SR)]和AIAS的差异。绘制受试者工作特征(ROC)曲线分析SR、AIAS单独及联合鉴别甲状腺良恶性结节的诊断效能。结果甲状腺恶性结节SR和AIAS均明显高于良性结节,差异均有统计学意义(均P<0.05)。ROC曲线分析显示,SR、AIAS鉴别诊断甲状腺良恶性结节的曲线下面积分别为0.885、0.845,灵敏度分别为75.00%、78.12%,特异度分别为97.78%、93.33%;二者联合鉴别诊断甲状腺良恶性结节的曲线下面积为0.917,灵敏度为82.81%,特异度为91.11%,均高于其单独应用,差异均有统计学意义(均P<0.05)。结论AIAS联合SWE在鉴别诊断甲状腺良恶性结节中有一定的临床价值。

Objective To explore the clinical value of artificial intelligence assisted score(AIAS)combined with shear wave elastography(SWE)in the differential diagnosis of benign and malignant thyroid nodules.Methods A total of 231 patients with thyroid nodules(327 nodules)confirmed by surgery or needle biopsy in our hospital were enrolled,including 135 benign nodules and 192 malignant nodules.The differences of SWE parameter[ratio of strain rate(SR)]and AIAS in benign and malignant thyroid nodules were compared.Receiver operating characteristic(ROC)curve was drawn to analyze the diagnostic efficacy of SR,AIAS alone and in combination in the differential diagnosis of benign and malignant thyroid nodules.Results SR and AIAS in malignant thyroid nodules were significantly higher than those in benign nodules(both P<0.05).ROC curve analysis showed that the area under the curve of SR and AIAS for differentiating benign and malignant thyroid nodules were 0.885 and 0.845,the sensitivity were 75.00%and 78.12%,and the specificity were 97.78%and 93.33%,respectively.The area under the curve,sensitivity and specificity of SR and AIAS combined for differentiating benign and malignant thyroid nodules were 0.917,82.81%and 91.11%,respectively,which were higher than those of individual application,and the differences were statistically significant(all P<0.05).Conclusion AIAS combined with SWE has certain clinical value for differentiating benign and malignant thyroid nodules.