甲状腺影像报告与数据系统(TI-RADS)分类鉴别良恶性甲状腺结节的临床应用价值

摘要 目的：探讨甲状腺影像报告与数据系统(TI-RADS)分类鉴别良恶性甲状腺结节(TN)的临床应用价值。方法：选取2016年5月至2019年12月我院收治的182例TN患者的临床资料进行回顾性分析，所有患者均完成超声检查和TI-RADS分类，以病理结果为金标准，比较良恶性结节超声征象及TI-RADS分布情况，并采用ROC曲线分析TI-RADS分类对TN良恶性的鉴别价值。结果：182例患者247个甲状腺结节中，良性172个，占比69.64%，恶性75个，占比30.36%；良性TN中实性结节、极低回声、边界模糊、微钙化和纵横比<1发生率以及结节直径均低于恶性TN，差异有统计学意义(P<0.05)；良性TN的TI-RADS分类主要为1类（59.30%）和2类（31.40%），恶性TN主要为3类（42.67%）和4类（46.67%），两组分布比较差异有统计学意义(P<0.05)；ROC曲线分析结果显示TI-RADS分类鉴别良恶性TN的AUC=0.939(95%CI：0.904~0.973，P<0.05)，以TI-RADS 2类和3类为良性，4类和5类为恶性进行诊断，其灵敏度为0.893，特异度为0.907，准确率为0.903，一致性Kappa值为0.777。结论：TI-RADS分类鉴别良恶性甲状腺结节具有较高的临床应用价值，但也还存在一定误诊的可能性。

Clinical Application Value of Thyroid Imaging Reporting and Data System (TI-RADS) Classification in the Differential Diagnosis of Benign and Malignant Thyroid Nodules

ABSTRACT Objective: To explore the clinical application value of thyroid imaging reporting and data system (TI-RADS) classification in the differential diagnosis of benign and malignant thyroid nodules (TN). Methods: The clinical data of 182 patients with TN treated in our hospital from May 2016 to December 2019 were retrospectively analyzed. All patients completed ultrasound and TI-RADS classification. The pathological results were used as the gold standards, and the ultrasound signs of benign and malignant nodules and TI-RADS distribution were compared, and ROC curve analysis was used to analyze the differential value of TI-RADS classification on benign and malignant TN. Results: Among the 247 thyroid nodules in 182 patients, 172 nodules were benign (accounting for 69.64%) and 75 nodules were malignant (accounting for 30.36%). The incidence rates of solid nodules, very low echo, blurred borders, microcalcification and aspect ratio<1 and nodule diameter of benign TN were lower than those of malignant TN(P<0.05). TI-RADS classifications were mainly Class 1 (59.30%) and Class 2 (31.40%) in benign TN, and were mainly Class 3 (42.67%) and Class 4 (46.67%) in malignant TN, and there was a statistically significant difference in the distribution between the two groups(P<0.05). ROC curve analysis results showed that AUC of TI-RADS classification in the identification of benign and malignant TN was 0.939 (95% CI: 0.904~0.973, P<0.05), and TI -RADS Class 2 and Class 3 were taken as benign and Class 4 and Class 5 were taken as malignant for diagnosis, and the sensitivity, specificity, accuracy rate and consistency Kappa were 0.893, 0.907, 0.903 and 0.777. Conclusion: TI-RADS classification has high clinical application value in the differentiation of benign and malignant thyroid nodules. However, there are still some misdiagnosis possibilities.