【摘要】 目的 基于实时二维剪切波弹性成像(2D SWE)构建慢性病毒性肝炎小肝癌(CVHSLC)预测模型。方法 收集病灶结节直径1～3 cm的慢性病毒性肝炎患者资料，按照病理诊断及增强影像学检查结果将结节分为恶性组55例和良性组85例，比较两组观察指标，行多因素Logistic回归分析影响小肝癌诊断的因素，并构建回归方程。利用受试者工作特征(ROC)曲线分析预测模型并预判其预测小肝癌的效能。结果 2D SWE测值是预测小肝癌的独立因素(P<0.05),建立CVHSLC预测小肝癌的Logistic回归模型，行Hosmer-Lemeshow检验提示模型有良好的校准度(χ~2=3.354,df=8,P=0.910)。利用ROC曲线检验该模型，与单独应用2D SWE相比，该模型预测小肝癌的曲线下面积(area under the curve, AUC)值具有显著性差异(Z=1.987,P=0.047)。结论 应用2D SWE构建慢性病毒性肝炎患者小肝癌预测模型的效能良好，可作为评估小肝癌的指标，具有很好的临床应用价值。更多还原

【Abstract】 Objective To construct a predictive model for chronic viral hepatitis small liver cancer(CVHSLC) based on real time two-dimensional shear wave elastography(2D SWE). Methods We collected data from patients diagnosed with chronic viral hepatitis with lesion nodules with a diameter of 1-3 cm.According to the pathological diagnosis and enhanced imaging examination results, the nodules were divided into the malignant group of 55 cases and the benign group of 85 cases. We compared the indicators of the two groups, performed multiple Logistic regression analysis on the factors affecting the diagnosis of small liver cancer, and constructed a regression equation. Receiver operating characteristic(ROC) curve was used to analyze the predictive model and its predictive value for the small liver cancer. Results 2D SWE values were independent factors for predicting small liver cancer(P<0.05). A Logistic regression model for predicting small liver cancer using CVHSLC was established, and a Hosmer Lemeshow test was performed, indicating good calibration of the model(χ~2=3.354,df=8,P=0.910). The ROC curve was used to test the model, and compared with using 2D SWE alone, the model showed a significant difference AUC value in predicting the small liver cancer(Z=1.987, P=0.047). Conclusions The application of 2D SWE to construct a prediction model for small liver cancer in patients with chronic viral hepatitis has good predictive efficacy and good clinical application value.更多还原