期刊文献+

超声诊断肝细胞肝癌术后5年复发情况的预测指数模型

PI model of ultrasonic diagnosis predicting the 5-year recurrence of postoperative hepatocellular carcinoma
下载PDF
导出
摘要 目的探讨影响肝细胞肝癌(HCC)术后5年复发情况的超声诊断因素并建立预测指数(PI)模型,评价该模型的预测价值。方法对所有相关超声诊断因素进行Kaplan-Meier和Log-rank单因素分析和多因素Cox回归分析,得出预测因子以及回归系数,根据预测因子以及回归系数建立超声诊断预测HCC术后5年复发情况的PI模型,对PI模型进行有效性检验和多因素校正,绘制PI模型预测HCC术后5年复发情况的ROC曲线,评价该模型对于判断HCC术后5年复发情况的预测价值。结果多因素Cox回归分析结果示:肿瘤边界(β=0.14,OR=1.23)、肿瘤部位(β=0.36,OR=1.58)、肿瘤大小(β=0.59,OR=2.42)、肿瘤数目(β=0.41,OR=1.79)为影响HCC术后5年复发情况的独立预测因子,根据预测因子以及回归系数建立PI预测模型:PI=0.14×X1(边界清=0,边界不清=1)+0.36×X2(周围型=0,中央型=1)+0.59×X3(cm)+0.41×X4(单发=0,多发=1),PI的临界值为1.66,模型有效性(χ^(2)=12.43,P<0.05),PI模型经多因素校正成为独立预测因子(β=1.08,OR=2.91),PI模型的ROC曲线下的面积为0.812(0.765~0.957),准确度为78.7%。结论本研究提出了超声诊断预测HCC术后5年复发情况的PI模型以及预测临界值,该模型的预测价值较高,为临床对HCC术后复发情况的精准评估提供的依据。 Objective To explore the ultrasonic diagnosis factors influencing the 5-year recurrence of postoperative hepatocellular carcinoma(HCC)and establish Predict Index(PI)model.Methods Kaplan-Meier and Log-rank univariate analysis and multivariate Cox regression analysis were used to analyze all related ultrasonic diagnosis factors,and the predictive factors and regression coefficients were obtained.According to the predictive factors and regression coefficients,a PI model was established to predict the 5-year recurrence of postoperative HCC by ultrasonic diagnosis.The validity of the PI model was tested and multivariate adjusted to draw ROC curve of PI model to predict the 5-year recurrence of postoperative HCC and evaluate the predictive value of this model for the 5-year recurrence of postoperative HCC.Results Multivariate Cox regression analysis showed that tumor boundary(β=0.14,OR=1.23),tumor site(β=0.36,OR=1.58),tumor size(β=0.59,OR=2.42),tumor number(β=0.41,OR=1.79)were independent predictors of the 5-year recurrence of postoperative HCC.According to the predictors and regression coefficient,PI prediction model was established as:PI=0.14×X1(clear boundary=0,unclear boundary=1)+0.36×X2(surrounding type=0,central type=1)+0.59×X3(cm)+0.41×X4(single engine=0,multiple engine=1),the cut off of PI was 1.66,the validity of PI was 12.43 byχ^(2) test(P<0.05).PI model was an independent predictor after multivariate factor correction(β=1.08,OR=2.91),the area under ROC curve of PI model was 0.812(0.765-0.957),and the accuracy was 78.7%.Conclusion We put forward PI model of ultrasonic diagnosis predicting the 5-year recurrence of postoperative HCC and the critical value at the first time.This model had a high predictive value,which provided a strong basis with clinical accurate assessment of the recurrence of postoperative HCC.
作者 阿曼·玛克沙提 庄国梅 杨小萍 邹媛 刘涛 王淑霞 AMAN Maksati;ZHUANG Guomei;YANG Xiaoping;ZOU Yuan;LIU Tao;WANG Shuxia(Department of Ultrasound,Health Management Institute of Xinjiang Medical University,Urumqi 830011,China)
出处 《分子影像学杂志》 2021年第1期8-12,共5页 Journal of Molecular Imaging
基金 国家自然科学基金(81860510)。
关键词 肝细胞肝癌 超声诊断 复发 PI模型 hepatocellular carcinoma ultrasonic diagnosis recurrence PI model
  • 相关文献

参考文献8

二级参考文献22

共引文献116

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部