Logistic回归分析多层螺旋CT在甲状腺良恶性结节诊断中的价值

目的 评价和筛选多层螺旋CT(MSCT)在甲状腺结节良恶性鉴别诊断中的特征指标。方法 回顾性分析经病理证实的105例甲状腺结节患者143个结节的MSCT资料,应用Logistic回归模型,筛选甲状腺良恶性结节鉴别的MSCT影像学特征。结果 甲状腺良、恶性结节在数目、形态、边界、钙化、囊变、强化特点、边缘中断征及淋巴结肿大等征象上有统计学差异;在结节部位、最大直径上无统计学差异。Logistic回归分析:边界、钙化、强化特点有统计学差异。结论 边界、钙化、强化特点是甲状腺良恶性结节鉴别的特征性预测指标。

Logistic regression analysis about the diagnostic value of multisliecs helieal CT on thyroid benign and malignant nodules

Objective To evaluate and screen out the characters of multisliecs helieal CT (MSCT) on thyroid benign and malignant nodules. Methods 143 thyroid nodules of the 105 patients were reviewed retrospectively with MSCT. By using the Logistic model, the radiological features by dependent MSCT were screened out. Results There were significant differences in the number, morphology, border, calcification, cysts, enhancement characteristics, marginal discontinuation and lymph node enlargement of thyroid benign and malignant nodules on MSCT. No significant difference in the maximum diameter and nodule position, Logistic regression analysis showed statistically significant differences in boundary, calcification and enhancement. Conclusion Boundary, calcification and enhancement of the thyroid nodules were characteristic predictors for the differential diagnosis of benign and malignant thyroid nodules.