基于超声造影组学特征构建甲状腺乳头状癌消融术后复发风险的列线图模型  

Construction of a columnar line graph model for the risk of recurrence after ablation of papillary thyroid cancer based on ultrasonographic histological features

作　　者：陈姗姗 闫静茹 何宁 韩增辉 CHEN Shan-shan;YAN Jing-ru;HE Ning;HAN Zeng-hui(Ultrasonic Diagnosis and Treatment Center,Xi'an International Medical Center Hospital,Xi'an 710100,Shaanxi,CHINA)

机构地区：[1]西安国际医学中心医院超声诊疗中心,陕西西安710100

出　　处：《海南医学》2023年第19期2840-2845,共6页Hainan Medical Journal

基　　金：陕西省西安市科学技术局项目[编号:20YXYJ0010(6)]。

摘　　要：目的探讨超声造影组学预测甲状腺乳头状癌(PTC)消融术后复发风险的价值。方法前瞻性选取2019年10月至2021年10月西安国际医学中心医院收治的100例行微波消融术的PTC患者,根据术后1年复发情况分为复发组(n=21)和未复发组(n=79)。比较两组患者的一般资料和超声造影组学特征,采用COX回归方程筛选PTC消融术后复发的影响因素,构建列线图模型,采用受试者工作曲线(ROC)评估模型预测价值并验证。结果两组患者的肿瘤分期、肿瘤直径比较差异均有统计学意义(P<0.05);复发组患者的PEAK、SI_(max)、SI_(mean)分别为(30.12±3.34)dB、(39.95±4.03)dB、(31.24±3.18)dB,明显低于未复发组的(37.42±4.05)dB、(47.56±4.33)dB、(38.20±4.19)dB,差异均有统计学意义(P<0.05);经COX回归方程分析结果显示,肿瘤分期、肿瘤直径、PEAK、SI_(max)、SI_(mean)均是PTC消融术后复发的影响因素(P<0.05);列线图预测模型显示,各预测指标对应分数值预测PTC消融术后复发的发生率为63%;经ROC分析结果显示,训练集患者AUC为0.844,C-index为0.844;验证集患者AUC为0.839,C-index为0.839;预测模型在训练集和验证集中GiViTI校准曲线带的80%~95%置信区间区域均未穿过45°角平分线,差异无统计学意义(P>0.05)。结论肿瘤分期、肿瘤直径、PEAK、SI_(max)、SI_(mean)均是PTC消融术后复发的影响因素,基于PEAK、SI_(max)、SI_(mean)等因素构建的列线图风险模型可有效预测消融术后复发风险。Objective To establish a line graph model to investigate the value of ultrasonography histology to predict the risk of recurrence after ablation of papillary thyroid cancer(PTC).Methods One hundred patients with PTC admitted to Xi'an International Medical Center Hospital for microwave ablation from October 2019 to October 2021 were prospectively selected and divided into a recurrence group(n=21)and a non-recurrence group(n=79)based on recurrence at 1 year after surgery.The general data and ultrasonographic histological characteristics of the two groups were compared,and the COX regression equation was used to screen the factors influencing recurrence after PTC ablation.A columnar line graph model was constructed,and the predictive value of the model was assessed and validated us-ing the subject operating curve(ROC).Results The differences in tumor stage and tumor diameter between the two groups were statistically significant(P<0.05).The PEAK,SI_(max),and SI_(mean) of patients in the recurrence group were(30.12±3.34)dB,(39.95±4.03)dB,and(31.24±3.18)dB,respectively,which were significantly lower than(37.42±4.05)dB,(47.56±4.33)dB,(38.20±4.19)dB in the non-recurrence group(P<0.05).The results of COX regression equation analysis showed that tumour stage,tumour diameter,PEAK,SI_(max),and SI_(mean) were all influencing factors for recurrence after PTC ablation(P<0.05).The prediction model showed that the corresponding score predicted 63%of recurrence after PTC ablation.ROC analysis showed that the AUC was 0.844 and the C-index was 0.844 in the training set,and 0.839 and 0.839 in the validation set.The prediction model showed that the 80%-95%confidence interval region of the GiViTI calibration curve band did not cross the 45°angle bisector in both the training and validation sets,and the difference was not statistically significant(P>0.05).Conclusion Tumor stage,tumor diameter,PEAK,SI_(max),SI_(mean) are the influencing factors of recurrence after PTC ablation,and the column line graph risk model constructed based

关 键 词：甲状腺乳头状癌 消融术 超声造影 复发因素 

分 类 号：R736.1[医药卫生—肿瘤]