肝脏 ›› 2025, Vol. 30 ›› Issue (7): 912-916.

• 肝纤维化及肝硬化 • 上一篇    下一篇

基于超声剪切波弹性成像建立肝硬化患者食管胃底静脉曲张的危险因素及预测模型

杨千冬, 付文学, 徐林林   

  1. 246700 安徽安庆 安庆市立医院超声医学科(杨千冬,付文学),消化内科(徐林林)
  • 收稿日期:2024-11-07 出版日期:2025-07-31 发布日期:2025-08-11
  • 基金资助:
    安徽省自然科学基金资助项目 (2208085MH259)

The risk factors and the prediction model based on ultrasonic shear wave elastography for esophageal and gastric varices rupture in cirrhotic patients

YANG Qian-dong1, FU Wen-xue1, XU Lin-lin2   

  1. 1. Department of Ultrasound Medicine , Anqing Municipal Hospital, Anqing 246700, China;
    2. Department of Gastroenterdogy, Anqing Municipal Hospital, Anqing 246700, China
  • Received:2024-11-07 Online:2025-07-31 Published:2025-08-11

摘要: 目的 基于超声剪切波弹性成像评估肝硬化患者食管胃底静脉曲张(EGV)的危险因素,构建并验证预测模型。方法 选取安庆市立医院2021年8月—2024年9月收治的98例肝硬化患者,依据8∶2定量随机分为训练集(n=78)与验证集(n=20)。分析肝硬化患者发生EGV的因素,构建并验证预测模型。结果 训练集中,有22例患者发生EGV。验证集中,有6例患者发生EGV。训练集中,发生组与未发生组在 Child-Pugh 分级(为B、C级)、脾大、脾脏硬度(SS)和肝脏硬度(LS)方面的比较,差异均有统计学意义(P<0.05)。经二元logistic回归分析,合并脾大[OR=4.173(95% CI:1.454~11.972)]、SS水平[OR=1.291(95% CI:1.118~1.492)]、LS水平[OR=1.662(95% CI:1.270~2.176)]、Child-Pugh分级为B、C级[OR=3.652(95% CI:1.184~11.265)]为肝硬化患者发生EGV的影响因素(P<0.05)。训练集模型预测肝硬化患者发生EGV的敏感度为90.91%,特异度为89.82%,曲线下面积为0.894;验证集模型预测肝硬化患者发生EGV的灵敏度为86.40%,特异度为84.50%,曲线下面积为0.879。结论 Child-Pugh分级越高、合并脾大、SS、LS与肝硬化患者发生EGV相关,建立的列线图预测模型对肝硬化患者发生EGV的预测效能良好。

关键词: 剪切波弹性成像, 肝硬化, 食管胃底静脉曲张, 危险因素, 预测模型

Abstract: Objective To analyze the risk factors of esophageal and gastric varices (EGV) in patients with cirrhosis based on ultrasonic shear wave elastography, and to construct and verify the prediction model. Methods Ninety-eight patients with cirrhosis admitted to Anqing Municipal Hospital from August 2021 to September 2024 were randomly divided into a training set (n=78) and a validation set (n=20) according to 8:2 quantitative analysis. The factors of EGV occurrence in cirrhotic patients were analyzed, and the predictive model was constructed and validated. Results In the training set, 22 patients developed EGV. In the validation set, EGV occurred in 6 cases. The Child-Pugh grades were B and C, combined with splenomegaly, SS and LS (P<0.05). Binary logistic regression analysis showed that splenomegaly [OR=4.173 (95%CI: 1.454~11.972)], SS level [OR=1.291 (95%CI: 1.118~1.492)], LS level [OR=1.662 (95%CI: 1.270~2.176)] and Child-Pugh grade B and C [OR=3.652 (95%CI: 1.184~11.265)] were the influencing factors for EGV occurrence in cirrhotic patients (P<0.05). The sensitivity and specificity of the training set model to predict EGV in patients with cirrhosis were 90.91%, 89.82%, and 0.894, respectively. The sensitivity and specificity of the validation set model for predicting EGV in patients with cirrhosis were 86.40%, 84.50%, and 0.879, respectively. Conclusion The incidence of EGV in cirrhosis patients is correlated with higher Child-Pugh grades, along with the presence of splenomegaly, SS, and LS. Developing a nomogram prediction model proves to be an effective tool for forecasting the likelihood of EGV occurrence in these patients.

Key words: Shear wave elastic imaging, Liver cirrhosis, Esophageal and gastric varices, Risk factors, Prediction model