超声结节-肌肉灰阶比值对甲状腺结节良恶性的预测价值研究

Value of nomogram model based on ultrasonographic parameters in differentiating benign and malignant thyroid nodules

目的探讨超声结节-肌肉灰阶比值对甲状腺结节良恶性的预测价值。方法回顾性选取2020年5月至2022年12月在杭州市余杭区第三人民医院及杭州市第一人民医院就诊的甲状腺结节患者152例(157个结节)。采用Adobe Photoshop CS6软件,对甲状腺结节、甲状腺组织和颈部肌肉组织的灰阶值进行测量并计算超声灰阶比值,同时提取结节的超声特征。采用多因素logistic回归分析结节良恶性的影响因素,基于此建立评估甲状腺结节良恶性的列线图模型,并采用ROC曲线分析其预测效能。结果157个结节中良性结节77个,恶性结节80个。多因素logistic回归分析结果显示超声下最大径(OR=0.823,95%CI:0.734~0.923)、结节-肌肉灰阶比值(OR=0.207,95%CI:0.057~0.749)、边缘情况(OR=0.298,95%CI:0.130~0.682)、纵横比(OR=0.177,95%CI:0.071~0.441)均是甲状腺结节良恶性的独立影响因素(均P<0.05)。纳入这4个超声特征,建立直观的评估甲状腺结节良恶性的列线图模型,ROC曲线分析显示AUC为0.845(95%CI:0.783~0.907),当约登指数最高时灵敏度和特异度分别为0.787、0.844,准确度为0.815。结论超声结节-肌肉超声灰阶比值可有效量化甲状腺结节回声强度。基于超声结节-肌肉灰阶比值、超声下结节最大径、纵横比、边缘情况构建的列线图模型对结节良恶性有较好的预测效能。

Objective To investigate the value of nomogram model based on ultrasonographic parameters in differentiating benign and malignant thyroid nodules.Methods A total of 152 patients with thyroid nodules admitted in Yuhang District Third People's Hospital and Hangzhou First People's Hospital from May 2020 to December 2022 were included in the study.Adobe Photoshop CS6 software was used to measure the gray scale values of thyroid nodules,thyroid tissue and muscle tissue on ultrasonographic images,and the ultrasonic gray scale ratios were calculated.Multivariate logistic regression was used to analyze the influencing factors of benign and malignant nodules,based on which a nomogram model was developed to differentiate benign and malignant thyroid nodules,and ROC curve was used to analyze its differentiating efficiency.Results A total of 157 thyroid nodules were detected in 152 patients,including 77 benign nodules and 80 malignant ones.Multivariate logistic regression analysis showed that the maximum diameter(OR=0.823,95%CI:0.734-0.923)and nodule-muscle gray scale ratio(OR=0.207,95%CI:0.057-0.749),edge case(OR=0.298,95%CI:0.130-0.682),aspect ratio(OR=0.177,95%CI:0.071-0.441)were independent factors influencing factors of benign and malignant thyroid nodules(all P<0.05).Incorporating these four ultrasonic features,a nomogram model was generated to differentiate benign and malignant thyroid nodules.The area under ROC curve(AUC)of the model in differentiating benign and malignant thyroid nodules was 0.845(95%CI:0.783-0.907),and the sensitivity and specificity were 0.787 and 0.844,respectively,and the accuracy was 0.815,when the Yodon index was the highest.Conclusion The nomogram model based on the ratio of ultrasonic nodule to muscle gray scale,the maximum diameter of the nodule,the aspect ratio and the edge of the nodule has a good prediction effect in differentiating the benign and malignant thyroid nodules.