
浏览全部资源
扫码关注微信
1.川北医学院临床医学院,南充 637000
2.中国人民解放军西部战区总医院消化内科,成都 610083
任 静,Email:renjing3036@163.com,研究方向:肝衰竭临床与基础研究。
汤善宏,Email:tangshanhong@swjtu.edu.cn。
收稿:2026-03-22,
修回:2026-06-05,
录用:2026-06-08,
纸质出版:2026-07-01
移动端阅览
任静, 柏承志, 苏绍辉, 等. 基于机器学习构建失代偿期乙型肝炎肝硬化并发ACLF患者的短期预后模型[J/OL]. 重庆医科大学学报, 2026,1-8.
Ren Jing,Bai Chengzhi,Su Shaohui,et al.Construction of a short-term prognostic model for patients with decompensated hepatitis B cirrhosis complicated by acute-on-chronic liver failure based on machine learning[J].Journal of Chongqing Medical University,
任静, 柏承志, 苏绍辉, 等. 基于机器学习构建失代偿期乙型肝炎肝硬化并发ACLF患者的短期预后模型[J/OL]. 重庆医科大学学报, 2026,1-8. DOI:
Ren Jing,Bai Chengzhi,Su Shaohui,et al.Construction of a short-term prognostic model for patients with decompensated hepatitis B cirrhosis complicated by acute-on-chronic liver failure based on machine learning[J].Journal of Chongqing Medical University, DOI:10.13406/j.cnki.cyxb..
目的
2
基于机器学习构建失代偿期乙型肝炎肝硬化并发慢加急性肝衰竭(acute-on-chronic liver failure,ACLF)患者的90 d预后模型,并评估其预测效能。
方法
2
回顾性收集2015年1月至2023年1月于西部战区总医院收治的218例失代偿期乙型肝炎肝硬化并发ACLF患者的临床资料,将患者以7∶3的比例随机分为训练集(153例)和验证集(65例),训练集用于构建模型,验证集用于验证模型;通过最小绝对收缩与选择算子(least absolute shrinkage and selection operator,LASSO)回归分析方法筛选与其预后相关的关键变量,采用Logistic回归、梯度提升机(Gradient Boosting Machine,GBM)、随机森林(Random Forest)、K近邻算法(K-nearest neighbor,KNN)、轻量级梯度提升机(Light Gradient Boosting Machine,LightGBM)等10种机器学习算法构建失代偿期乙型肝炎肝硬化并发ACLF患者90 d短期预测模型,绘制受试者工作特征曲线(receiver operating characteristic curve,ROC),比较ROC曲线下面积(area under the curve,AUC)、准确度、灵敏度、特异度和F1分数等指标评估不同模型的预测效能。通过沙普利可加性解释(Shapley additive explanations,SHAP)分析各关键变量对最优模型预测结果的贡献度并排序。
结果
2
通过LASSO回归筛选出6个变量,基于此构建10种机器学习模型,其中LightGBM模型预测效能最佳,AUC为0.813,准确度为0.769,SHAP图显示重要性排序分别为总胆红素(total bilirubin,TBIL)、国际标准化比值(international normalized ratio,INR)、肝性脑病(hepatic encephalopathy,HE)、白细胞(white blood cell count,WBC)、年龄(age)和低密度脂蛋白胆固醇(low-density lipoprotein cholesterol,LDL-C)。
结论
2
成功构建出预测失代偿期乙型肝炎肝硬化并发ACLF患者90 d预后的模型,其中LightGBM模型预测效能最佳,有助于识别高风险患者,为临床决策及干预提供依据。
Objective
2
To construct a predictive model for the 90-day prognosis of patients with decompensated hepatitis B cirrhosis complicated by acute-on-chronic liver failure(ACLF),and to assess its predictive efficacy.
Methods
2
A retrospective analysis was performed for the clinical data of 218 patients with decompensated hepatitis B cirrhosis complicated by ACLF who were admitted to The General Hospital of Western Theater Command from January 2015 to January 2023,and these patients were randomly divided into a training set with 153 patients and a validation set with 65 patients at a ratio of 7∶3,with the training set used for model construction and the validation set for model validation. The least absolute shrinkage and selection operator(LASSO) regression analysis was used to identify the key variables associated with prognosis,and 10 machine learning algorithms were used to construct a predictive model for the 90-day short-term prognosis of patients with decompensated hepatitis B cirrhosis complicated by ACLF,including logistic regression model,gradient boosting machine(GBM),random forest,k-nearest neighbor(KNN),and light gradient boosting machine (LightGBM). The receiver operating characteristic(ROC) curve was plotted,and the area under the ROC curve(AUC),accuracy,sensitivity,specificity,and F1-score of different models were compared to assess their predictive efficacy. Shapley Additive exPlanations(SHAP) was used to analyze and rank the contribution of key variables to the prediction results of the optimal model.
Results
2
Six variables were identified by LASSO regression,and 10 machine learning models were constructed based on these variables,among which the LightGBM model showed the optimal predictive performance,with an AUC of 0.813 and an accuracy of 0.769. The SHAP plots showed that total bilirubin had the greatest importance,followed by international normalized ratio,hepatic encephalopathy,white blood cell count,age,and low-density lipoprotein cholesterol.
Conclusion
2
This study successfully constructed the models for predicting the 90-day prognosis of patients with decompensated hepatitis B cirrhosis complicated by ACLF,among which the LightGBM model has the best predictive performance and can help to identify high-risk patients and provide a basis for clinical decision-making and intervention.
Polaris Observatory Collaborators . Global prevalence,cascade of care,and prophylaxis coverage of hepatitis B in 2022:a modelling study [J ] . Lancet Gastroenterol Hepatol , 2023 , 8 ( 10 ): 879 - 907 .
中华医学会感染病学分会肝衰竭与人工肝学组 , 中华医学会肝病学分会重型肝病与人工肝学组 . 肝衰竭诊治指南(2024年版) [J ] . 中华肝脏病杂志 , 2025 , 33 ( 01 ): 18 - 33 .
Liver Failure and Artificial Liver Group , Chinese Society of Infectious Diseases , Chinese Medical Association ; Severe Liver Disease and Artificial Liver Group , Chinese Society of Hepatology , Chinese Medical Association . Guideline for diagnosis and treatment of liver failure [J ] . Chin J Hepatol , 2025 , 33 ( 01 ): 18 - 33 .
Trebicka J , Praktiknjo M , Peiffer KH , et al . Acute on chronic liver failure [J ] . Dtsch Ärztebl Int , 2025 , 122 ( 4 ): 96 - 102 .
Li CY , Hu H , Bai CZ , et al . Alpha-fetoprotein and APRI as predictive markers for patients with Type C hepatitis B-related acute-on-chronic liver failure:a retrospective study [J ] . BMC Gastroenterol , 2024 , 24 ( 1 ): 191 .
Sarin SK , Choudhury A , Sharma MK , et al . Acute-on-chronic liver failure:consensus recommendations of the Asian Pacific association for the study of the liver (APASL):an update [J ] . Hepatol Int , 2019 , 13 ( 4 ): 353 - 390 .
孔雪杰 , 李 江 , 李昊天 , 等 . 不同肝病基础的慢加急性肝衰竭患者临床特征和预后评分模型预测效能比较 [J ] . 实用肝脏病杂志 , 2021 , 24 ( 3 ): 387 - 390 .
Kong XJ , Li J , Li HT , et al . Clinical feature and efficacy of different prognostic scoring models in patients with acute-on-chronic liver failure with different underlying liver diseases [J ] . J Prac Hepatol , 2021 , 24 ( 3 ): 387 - 390 .
韩亚飞 , 汪 静 , 张添添 , 等 . 基于机器学习的肠球菌血流感染预后不良预测模型的构建与评估 [J ] . 中华医院感染学杂志 , 2026 , 36 ( 5 ): 790 - 795 .
Han YF , Wang J , Zhang TT , et al . Establishment of prediction model for adverse prognosis of patients with Enterococcus bloodstream infection based on machine learning and its predictive efficiency [J ] . Chin J Nosocomiol , 2026 , 36 ( 5 ): 790 - 795 .
Deng B , Bai CZ , Xu HQ , et al . Machine learning-based algorithms for the prediction of 90-day survival in patients with liver failure receiving artificial liver therapy [J ] . Front Physiol , 2025 , 16 : 1687860 .
中华医学会感染病学分会肝衰竭与人工肝学组 , 中华医学会肝病学分会重型肝病与人工肝学组 . 肝衰竭诊治指南(2018年版) [J ] . 中华肝脏病杂志 , 2019 , 27 ( 1 ): 18 - 26 .
Liver Failure and Artificial Liver Group , Chinese Society of Infectious Diseases , Chinese Medical Association ; Severe Liver Disease and Artificial Liver Group , Chinese Society of Hepatology , Chinese Medical Association . Guideline for diagnosis and treatment of liver failure [J ] . Chin J Hepatol , 2019 , 27 ( 1 ): 18 - 26 .
Li JQ , Liang X , You SL , et al . Development and validation of a new prognostic score for hepatitis B virus-related acute-on-chronic liver failure [J ] . J Hepatol , 2021 , 75 ( 5 ): 1104 - 1115 .
Ferenci P , Lockwood A , Mullen K , et al . Hepatic encephalopathy:Definition,nomenclature,diagnosis,and quantification:Final report of the Working Party at the 11th World Congresses of Gastroenterology,Vienna,1998 [J ] . Hepatology , 2002 , 35 ( 3 ): 716 - 721 .
张仕华 , 柏承志 , 李春燕 , 等 . 老年HBV相关慢加急性肝衰竭预后的危险因素及风险预测列线图模型构建 [J ] . 临床肝胆病杂志 , 2024 , 40 ( 10 ): 1976 - 1984 .
Zhang SH , Bai CZ , Li CY , et al . Risk factors for the prognosis of elderly patients with hepatitis B virus-related acute-on-chronic liver failure and construction of a nomogram model for risk prediction [J ] . J Clin Hepatol , 2024 , 40 ( 10 ): 1976 - 1984 .
Xu XW , Hou ZH , Xu YY , et al . The dynamic of platelet count as a novel and valuable predictor for 90-day survival of hepatitis B virus-related acute-on-chronic liver failure patients [J ] . Clin Res Hepatol Gastroenterol , 2021 , 45 ( 2 ): 101482 .
杜 蔷 , 陆盛菊 , 卢大秀 , 等 . HBV感染失代偿期肝硬化并发慢加急性肝衰竭患者的临床特点及预后 [J ] . 中国老年学杂志 , 2021 , 41 ( 3 ): 520 - 523 .
Du Q , Lu SJ , Lu DX , et al . Clinical characteristics and prognosis of patients with decompensated cirrhosis complicated with acute-on-chronic liver failure caused by HBV infection [J ] . Chin J Gerontol , 2021 , 41 ( 3 ): 520 - 523 .
牛 宏 , 姚 佳 , 崔少波 , 等 . 预后营养指数对人工肝治疗乙型肝炎相关慢加急性肝衰竭的预后评估价值 [J ] . 传染病信息 , 2025 , 38 ( 4 ): 320 - 325 .
Niu H , Yao J , Cui SB , et al . Prognostic value of the prognostic nutritional index in evaluating outcomes of artificial liver support system treatment for hepatitis B virus-related acute-on-chronic liver failure [J ] . Infect Dis Inf , 2025 , 38 ( 4 ): 320 - 325 .
邓珂欣 , 罗红春 . 常用血液学指标评定慢加急性肝衰竭患者并发细菌感染和预后价值研究 [J ] . 实用肝脏病杂志 , 2023 , 26 ( 2 ): 242 - 245 .
Deng KX , Luo HC . Evaluation of bacterial infection by common hematological indices in patients with acute-on-chronic liver failure [J ] . J Prac Hepatol , 2023 , 26 ( 2 ): 242 - 245 .
王 航 , 孙长峰 , 何 川 , 等 . 血脂水平对HBV相关慢加急性肝衰竭患者短期预后的预测价值 [J ] . 山东医药 , 2022 , 62 ( 8 ): 74 - 77 .
Wang H , Sun CF , He C , et al . Predictive value of blood lipid level on short-term prognosis in patients with HBV-related acute-on-chronic liver failure [J ] . Shandong Med J , 2022 , 62 ( 8 ): 74 - 77 .
官 敏 , 李嘉妮 , 高银杰 , 等 . 血脂指标对乙型肝炎相关慢加急性肝衰竭前期进展的预测价值 [J ] . 临床肝胆病杂志 , 2023 , 39 ( 7 ): 1564 - 1569 .
Guan M , Li JN , Gao YJ , et al . Value of blood lipid parameters in predicting the progression of HBV-related acute-on-chronic pre-liver failure [J ] . J Clin Hepatol , 2023 , 39 ( 7 ): 1564 - 1569 .
Xiao CC , Gong J , Zhu SG , et al . Nomogram based on blood lipoprotein for estimation of mortality in patients with hepatitis B virus-related acute-on-chronic liver failure [J ] . BMC Gastroenterol , 2020 , 20 ( 1 ): 188 .
Song BL , Liu P , Fu KR , et al . Developing a predictive model for septic shock risk in acute pancreatitis patients using interpretable machine learning algorithms [J ] . Digit Health , 2025 , 11 : 20552076251346361 .
Zhang JY , Lei YT , Liu R , et al . Development and validation of a machine learning-based explainable predictive model for long-term net adverse clinical events in patients with high bleeding risk undergoing percutaneous coronary intervention:results from a prospective cohort study [J ] . Int J Surg , 2025 , 111 ( 9 ): 6082 - 6092 .
Lynam AL , Dennis JM , Owen KR , et al . Logistic regression has similar performance to optimised machine learning algorithms in a clinical setting:application to the discrimination between type 1 and type 2 diabetes in young adults [J ] . Diagn Progn Res , 2020 , 4 ( 1 ): 6 .
0
浏览量
0
下载量
0
CSCD
关联资源
相关文章
相关作者
相关机构
京公网安备11010802024621