基于能谱CT参数列线图术前预测胃癌淋巴结转移的价值  

Clinical value of dual energy CT basedparameter nomogram in preoperative predictinglymph node metastasis of gastric cancer

作　　者：赵德雷 徐敏[2] 谢宗玉[2] Zhao Delei;Xu Min;Xie Zongyu(Graduate School of Bengbu Medical College,Bengbu,Anhui233030,China;Department of Radiology,the fFrst Affiliated Hospital of Bengbu Medical College,Bengbu,Anhui233004,China)

机构地区：[1]蚌埠医学院研究生院,安徽蚌埠233030 [2]蚌埠医学院第一附属医院放射科,安徽蚌埠233004

出　　处：《齐齐哈尔医学院学报》2023年第15期1443-1448,共6页Journal of Qiqihar Medical University

基　　金：蚌埠医学院自然科学重点项目(2021byzd179)。

摘　　要：目的探讨基于能谱CT(DECT)参数列线图在术前预测胃癌(GC)淋巴结转移中的应用价值。方法回顾性分析2020—2021年本院经手术治疗的180例GC患者的术前临床资料及DECT图像,以术后病理结果为金标准,结果显示,180例患者中115例存在淋巴结转移及65例未发现淋巴结转移。通过GSI viewer软件平台测量能谱参数,并进行CT征象评价。利用单因素及二元Logisitic回归分析对患者的临床资料、CT征象及能谱参数各项变量进行检验,以P<0.05的临床资料及CT征象变量构建临床预测模型,以P<0.05的能谱参数变量构建能谱预测模型,后将二者联合构建联合预测模型,并绘制列线图将联合预测模型可视化。采用受试者工作特征曲线(ROC)的曲线下面积(AUC)对三种预测模型的诊断效能进行评价。结果临床预测模型纳入3个变量:CEA、肿瘤厚度及脂肪浸润,能谱预测模型纳入1个变量:静脉期nIC;联合预测模型纳入4个变量:CEA、肿瘤厚度、脂肪浸润及静脉期nIC。三种预测模型的AUC值分别为0.773、0.772和0.856,敏感度分别为77.8%、70.0%、85.2%,特异度分别为80.0%、81.5%、83.8%,结果显示联合预测模型的诊断效能最高。同时联合预测模型的校准曲线表明对于GC淋巴结转移的预测概率与实际概率之间具有良好一致性。联合预测模型的Hosmer-Lemeshow检验拟合优度的P值为0.935(P>0.05)及诊断阈值为0.648。结论基于DECT参数列线图能够在术前更精准地预测GC患者的淋巴结状态,为临床个性化治疗提供更多依据,具有较好的临床应用价值。Objective To investigate the application value of nomogram based on parameter of dual energy CT(DECT)in preoperative prediction of lymph node metastasis(LNM)of gastric cancer(GC).Methods The preoperative clinical data and DECT images of 180 patients with GC who were treated surgically in our hospital from 2020 to 2021 were collected and analyzed retrospectively.According to the postoperative pathological results as the gold standard,the results showed that 115 patients had lymph node metastasis and 65 patients did not have lymph node metastasis.GSI viewer software was used to measurethe DECT parameters of the patients and the CT imaging features were evaluated at the same time.Using univariate and binary logistic regression analysis to screen the clinical data,CT imaging features and the parameters of DECT,the clinical data and CT imaging features with significant difference(P<0.05)were enrolled in the construction of clinical prediction model,the DECT parameter with significant difference(P<0.05)were used to establish dual energy prediction model.Then the two were combined to build a joint prediction model,and drew a nomogram to visualize the joint prediction model.The diagnostic efficacy of the three prediction models was evaluated using the area under the curve(AUC)of the receiver operating characteristic curve(ROC).Results The clinical prediction model included three variables:CEA,tumor thickness and fat infiltration,and DECT parameter prediction model included one variable:nIC of venous phase;The united prediction model included four variables:CEA,tumor thickness,fat infiltration and nIC of venous phase.The AUC values of the three prediction models were 0.773,0.772 and 0.856,respectively;the sensitivity were 77.8%,70.0%and 85.2%,respectively;the specificity was 80.0%,81.5%and 83.8%,respectively;the results indicated that the combined prediction model had the highest diagnostic efficiency.At the same time,the calibration curve of the joint prediction model showed that the prediction probability was in good ag

关 键 词：胃癌 淋巴结转移 能谱成像 

分 类 号：R735.2[医药卫生—肿瘤]