建立甲状腺微小结节的超声影像报告与数据系统（TI-RADS）分类。

2013年1月至2015年6月在上海交通大学医学院附属瑞金医院行超声检查并进行手术的529例甲状腺结节患者570个≤1 cm的甲状腺结节被纳入研究。（1）回顾性分析甲状腺结节的超声声像图特征；（2）通过Logistic回归分析，得到回归方程，并计算P^us值，推测结节存在不同恶性特征个数的情况下其恶性的发生概率；（3）将所有甲状腺结节根据恶性特征的个数分类，计算每一类中恶性结节占该类结节总数的比例，即实际恶性率；（4）根据上述结果，建立TI-RADS。

（1）共筛选出5个指标与甲状腺恶性结节独立相关，分别是边缘不光整、实性结构、极低回声、微钙化、边缘为主的血供。（2）随着恶性指标个数的增加，该甲状腺结节的恶性风险越大。（3）TI-RADS建立如下：TI-RADS 3类（恶性率≤5%），建议1年/次超声随访；TI-RADS 4a类（恶性率5%~20%），建议3个月/次的超声随访或穿刺；TI-RADS 4b类（恶性率20%~50%）及4c类（恶性率50%~90%），建议进行相应的穿刺或手术治疗；TI-RADS 5类（恶性率>90%），建议手术。

本研究所建立的TI-RADS系统具有较高的临床应用价值。甲状腺微小结节的恶性特征个数越多，其恶性风险越高。

To establish the imaging and data system of thyroid micronodule (TI-RADS).

From January 2013 to June 2015, a total of 529 patients with 570 thyroid micronodules (≤1 cm) underwent ultrasound examination and surgical treatment in Shanghai Rui Jin Hospital, Shanghai Jiao Tong University, School of Medicine. Firstly, all nodules were retrospectively analyzed on their ultrasonographic characteristics. Secondly, a regression equation was obtained via logistic regression analysis to calculate P^us values and to evaluate the probability of malignancy related to a certain number of suspected malignant characteristics. Thirdly, all the nodules were categorized according to how many suspected malignant characteristics they have. The actual proportion of malignant nodules was calculated, which was the actual malignancy rate. At last, TI-RADS was established on the basis of the results above.

(1) Five characteristics were independently associated with malignancy according to statistic analyses, which were non-smooth margin, solid structure, very low echo, microcalcification and blood supply mainly on the margin, respectively. (2) As the number of suspected malignant characteristics increased, the P^us value and the actual malignancy rate were higher, which indicated higher risk of malignancy. (3) TI-RADS is as follows: TI-RADS 3 (malignancy rate ≤5%), a follow up ultrasound examination once a year is recommended; TI-RADS 4a (malignancy rate 5%-20%), a follow up ultrasound examination every three months or fine needle aspiration is recommended; TI-RADS 4b (malignancy rate 20%-50%) and 4c (malignancy rate 50%-90%), fine needle aspiration or surgery is recommended; TI-RADS 5 (malignancy rate>90%), surgery is recommended.

TI-RADS is of great value in the clinical use. The more suspected malignant characteristics a micronodule has, the higher the risk of malignancy of the nodule is.