开发和验证急性至慢性肝衰竭的预后模型
Xia Zhu1,2, Ming Wang1,2, Yuanji Ma1,2
1Center of Infectious Diseases, West China Hospital of Sichuan University, Chengdu, China.
Frontiers in cellular and infection microbiology
|February 25, 2026
概括
一个新的机器学习模型准确地预测了B型肝炎病毒相关的急性至慢性肝衰竭 (HBV-ACLF) 的短期死亡率. 这种包含肝脏储备功能的模型,优于传统的评分系统,可以更好地管理患者.
科学领域:
- 肝病学 肝病学是一种肝病学.
- 机器学习在医学中的应用
- 预测模型的预测建模
背景情况:
- 对急性至慢性肝衰竭 (ACLF) 的预后评估,特别是在B型肝炎病毒 (HBV) 流行地区,具有挑战性.
- 传统模型在预测HBV-ACLF患者的短期结果方面显示出有限的准确性.
研究的目的:
- 开发和验证一种新的机器学习模型,以改善对HBV-ACLF短期结果的个性化预测.
- 将肝储备功能的指标纳入预后模型.
主要方法:
- 追溯收集了496名HBV-ACLF患者的数据,用于培训和52名外部验证.
- 对12个机器学习算法的系统评估,使用LASSO-RF方法选择最佳模型.
- 使用SHAP值识别关键预测变量,并与MELD和CTP得分进行比较.
主要成果:
- 最终的LASSO-RF模型实现了高预测准确性,在训练队列中AUC为0.99,在90天死亡率的验证队列中为0.98.
- 该模型显著超过了MELD和CTP等既有评分系统.
- 为临床应用开发了一个交互式网络计算器,提供实时风险得分.
结论:
- 肝储备功能指标对于预测HBV-ACLF结果至关重要.
- 开发的机器学习模型和在线工具为HBV-ACLF患者提供了准确的风险分层.
- 该工具有助于及时和个性化的临床管理决策.
相关概念视频
01:22Effect of Hepatic Disease on Pharmacokinetics: Pathophysiologic Assessment and Liver Function Test
244
In clinical practice, the direct measurement of hepatic blood flow to evaluate liver function presents significant challenges due to the intricate and specialized nature of the necessary techniques. Consequently, healthcare professionals often rely on empirical estimates derived from thorough patient examinations and liver function tests to gauge liver health. Among the tools at their disposal, the Child–Pugh and MELD scoring systems stand out for their ability to categorize and assess...
244
01:26Effect of Hepatic Disease on Pharmacokinetics: Drug Dosing and Hepatic Blood Flow
298
Chronic liver disease significantly impacts drug metabolism due to alterations in hepatic blood flow and enzyme accessibility. This disruption affects the body's pharmacokinetics—the movement and processing of drugs within the system. Key enzymes crucial for metabolizing medications become less accessible, changing how drugs are processed and utilized. Furthermore, liver disease influences the synthesis of plasma proteins, such as albumin and globulins, which play critical roles in drug...
298