甲状腺微小乳头状癌彩超征象危险因素的Logistic回归分析  

Logistic regression analysis of risk factors related to ultrasound diagnosis of papillary thyroid microcarcinoma

作　　者：张歆[1] 薛玉[1] 沈楚[1] 许建萍[1] 杜睿[1] 王玲[1] 

机构地区：[1]江苏大学附属医院超声医学科,江苏镇江212001

出　　处：《江苏大学学报（医学版）》2013年第1期75-77,共3页Journal of Jiangsu University：Medicine Edition

摘　　要：目的:探讨甲状腺微小癌的超声诊断危险因素与临床价值。方法:以手术病理结果为金标准,对术前行超声检查的123个甲状腺微小结节患者进行Logistic回归分析,筛选出与甲状腺微小癌相关度较高的超声特征参数,并建立Logistic回归方程。结果:Logistic回归分析显示最后进入方程模型的4个参数分别为形态不规则、边界模糊、内部低回声、内部微小钙化,它们与甲状腺微小癌有较高相关性(P<0.05)。结论:以形态、边界、内部钙化与结节内部回声4个诊断参数变量建立的Logistic回归模型有助于甲状腺微小癌的早期诊断。Objective： To evaluate the risk factors and the clinical value related to ultrasound diagnosis of papillary thyroid microcarcinoma（PTMC）.Methods： Taking pathological results as the gold standard,123 cases of small thyroid nodules diagnosed by ultrasound before surgery were analyzed by Logistic regression;a Logistic model would be obtained through making a screening test to find which ultrasound diagnosis parameters have a high relativity with PTMC.Results： The Logistic regression analysis demonstrated that 4 parameters were enrolled in the Logistic equation： irregular shape,unclear edge,microcalcifications,hypoechoic nodules;those four had a higher relativity with PTMC than the rest（P0.05）.Conclusion The Logistic regression model with shape,edge,microcalcifications,internal echoes can be helpful for early accurate diagnosis of PTMC.

关 键 词：甲状腺微小乳头状癌 超声 LOGISTIC回归分析 

分 类 号：R445.1[医药卫生—影像医学与核医学] R736.1[医药卫生—诊断学]