The School of Medicine, Nankai University, Tianjin 300071, China; Department of Gastroenterology and Hepatology, Tianjin Union Medical Center Affiliated to Nankai University, Tianjin 300121, China.
The School of Medicine, Nankai University, Tianjin 300071, China.
Dig Liver Dis. 2024 Dec;56(12):2095-2102. doi: 10.1016/j.dld.2024.06.029. Epub 2024 Jul 14.
We aimed to establish a prognostic predictive model based on machine learning (ML) methods to predict the 28-day mortality of acute-on-chronic liver failure (ACLF) patients, and to evaluate treatment effectiveness.
ACLF patients from six tertiary hospitals were included for analysis. Features for ML models' development were selected by LASSO regression. Models' performance was evaluated by area under the curve (AUC) and accuracy. Shapley additive explanation was used to explain the ML model.
Of the 736 included patients, 587 were assigned to a training set and 149 to an external validation set. Features selected included age, hepatic encephalopathy, total bilirubin, PTA, and creatinine. The eXtreme Gradient Boosting (XGB) model outperformed other ML models in the prognostic prediction of ACLF patients, with the highest AUC and accuracy. Delong's test demonstrated that the XGB model outperformed Child-Pugh score, MELD score, CLIF-SOFA, CLIF-C OF, and CLIF-C ACLF. Sequential assessments at baseline, day 3, day 7, and day 14 improved the predictive performance of the XGB-ML model and can help clinicians evaluate the effectiveness of medical treatment.
We established an XGB-ML model to predict the 28-day mortality of ACLF patients as well as to evaluate the treatment effectiveness.
本研究旨在建立一种基于机器学习（ML）方法的预测模型，以预测慢加急性肝衰竭（ACLF）患者 28 天病死率，并评估治疗效果。
纳入 6 家三级医院的 ACLF 患者进行分析。采用 LASSO 回归选择用于 ML 模型开发的特征。通过曲线下面积（AUC）和准确性评估模型的性能。使用 Shapley 加性解释法解释 ML 模型。
共纳入 736 例患者，其中 587 例患者被分配到训练集，149 例患者被分配到外部验证集。入选的特征包括年龄、肝性脑病、总胆红素、PTA 和肌酐。在 ACLF 患者预后预测方面，eXtreme Gradient Boosting（XGB）模型优于其他 ML 模型，具有最高的 AUC 和准确性。Delong 检验表明，XGB 模型优于 Child-Pugh 评分、MELD 评分、CLIF-SOFA、CLIF-C OF 和 CLIF-C ACLF。基线、第 3 天、第 7 天和第 14 天的连续评估提高了 XGB-ML 模型的预测性能，有助于临床医生评估治疗效果。
我们建立了一个 XGB-ML 模型，以预测 ACLF 患者 28 天病死率，并评估治疗效果。
