摘要
目的:探讨梯度提升机(GBM)模型在预测非静脉曲张上消化道出血(NVUDB)患者再出血中的临床价值。方法:回顾性分析2020年10月至2021年12月本院收治的258例NVUDB患者的临床资料,并按照7∶3比例将数据集随机分为训练集和验证集,分别用于构建GBM模型和验证模型的可靠性。采用受试者工作特征(ROC)曲线分析评价模型性能,校准曲线评估模型预测概率与样本概率之间的一致性,决策曲线评估该模型的临床实用性。结果:NVUDB患者再出血发生率为20.9%。GBM算法模型中重要特征得分前5项为Rockall评分、入院时休克、D-二聚体水平、白蛋白水平、红细胞分布宽度。训练集曲线下面积为0.985(95%CI:0.971~0.998),验证集为0.873(95%CI:0.785~0.960)。训练集的预测准确率为92.2%,验证集的预测准确率为83.3%。校准曲线显示GBM模型预测值与实际观测值之间具有良好的一致性,模型能够较好地预测实际概率。临床决策曲线分析结果展示了模型具有良好的临床表现能力。结论:基于GBM算法模型可以较好地预测NVUDB患者再出血的风险因素,且具有较高的临床有效性。
Objective:To explore the clinical value of gradient boosting machine(GBM)model in predicting rebleeding in patients with non-variceal upper digestive bleeding(NVUDB).Methods:Clinical data of 258 patients with NVUDB admitted to our hospital from October 2020 to December 2021 were retrospectively analyzed,and the data set was randomly divided into training set and validation set according to the ratio of 7∶3,which were used to construct GBM model and verify the reliability of the model,respectively.Receiver operating characteristic(ROC)curve was used to analyze and evaluate the performance of the model,the calibration curve was used to evaluate the consistency between the model prediction probability and the sample probability,and the decision curve was used to evaluate the clinical practicability of the model.Results:The incidence of rebleeding in NVUDB patients was 20.9%.The top five important feature scores in the GBM algorithm model were Rockall score,shock on admission,D-dimer level,albumin level,and red blood cell distribution width.The area under the curve(AUC)of the training set was 0.985(95%CI:0.971-0.998),and the validation set was 0.873(95%CI:0.785-0.960).The prediction accuracy of the training set was 92.2%,and the prediction accuracy of the validation set was 83.3%.The calibration curve showed that there was a good consistency between the predicted value of the GBM model and the actual observation value,and the model could predict the actual probability well.Decision curve analysis showed that the model had good clinical performance.Conclusion:The GBM algorithm model can better predict the risk factors of rebleeding in patients,and has high clinical effectiveness.
作者
吴江山
黄兴蔚
曾毅飞
莫兰梅
Wu Jiangshan;Huang Xingwei;Zeng Yifei;Mo Lanmei(Department of Gastroenterology,Wuming Hospital of Guangxi Medical University,Wuming 530199,China)
出处
《广西医科大学学报》
CAS
2023年第8期1334-1341,共8页
Journal of Guangxi Medical University
基金
广西壮族自治区卫生健康委员会自筹经费科研课题资助项目(No.Z20200077)。
关键词
机器学习
梯度提升机模型
非静脉曲张上消化道出血
再出血
决策曲线
machine learning
gradient boosting machine model
non-variceal upper digestive bleeding
rebleeding
decision curve
