文献速递心脑血管急症重症综合管理与镇静镇痛

人工智能早期预测颅内高压:从单中心开发到多中心验证

Artificial Intelligence for Early Prediction of Intracranial Hypertension: From Single-Centre Development to Multicentre Validation.
PubMed 文献 · Neurocrit Care · 2026-10-14
一句话结论基于Uppsala数据训练的XGBoost模型在CENTER-TBI多中心队列外部验证中保持高判别性能(AUROC 97.4%),但需本地评估后再临床实施。

该研究将Uppsala单中心高频TBI队列训练的XGBoost模型,用CENTER-TBI多中心高分辨率ICU队列进行双向外部验证。模型利用4小时历史窗口的分钟级ICP和有创动脉压数据,预测30分钟内颅内压 insult 的发生。Uppsala训练-CENTER-TBI测试的AUROC为97.4%,敏感性93.8%,特异性98.6%;反向验证AUROC为96.1%,敏感性81.2%,特异性98.7%。结果表明 insult 前生理模式跨队列共享,但敏感性差异提示需本地评估。

要点

  • 模型在CENTER-TBI外部验证中AUROC达97.4%,敏感性93.8%,特异性98.6%,支持跨中心泛化。
  • 反向验证(CENTER-TBI训练→Uppsala测试)AUROC 96.1%,敏感性81.2%,特异性98.7%,敏感性较低。
  • 预测目标为30分钟内ICP insult发生,基于4小时历史窗口的分钟级ICP和有创动脉压数据。
  • 双向验证平衡准确率均超过88%,但敏感性方向差异可能反映队列特征或数据集大小。
  • 结论强调临床实施前需进行本地评估,不能直接外推。

适用人群与场景

  • 纳入:Uppsala高频TBI队列602例和CENTER-TBI高分辨率ICU队列255例的TBI患者。
  • 场景:ICU内使用分钟级ICP和有创动脉压监测数据的TBI患者。
  • 排除:原文未提及具体排除标准。

诊断标准与关键阈值

  • 预测目标:基于4小时历史窗口的分钟级ICP和有创动脉压数据,预测30分钟内颅内压insult的发生。
  • 模型:XGBoost机器学习模型,识别与即将发生的阈值跨越相关的短期生理轨迹。

处置流程与药物剂量

  • 原文未提及

红旗征与不能漏的情况

  • 原文未提及具体红旗征或不能漏诊的情况及处理。

研究设计与关键结果

  • 研究类型:外部验证研究,使用Uppsala单中心队列训练模型,CENTER-TBI多中心队列进行验证。
  • 样本量与中心数:Uppsala 602例(单中心),CENTER-TBI 255例(多中心),具体中心数原文未提及。
  • 人群:TBI患者,使用分钟级ICP和有创动脉压数据。
  • 干预与对照:无干预,为模型预测性能验证。
  • 主要终点结果:Uppsala训练→CENTER-TBI测试AUROC 97.4%,敏感性93.8%,特异性98.6%;CENTER-TBI训练→Uppsala测试AUROC 96.1%,敏感性81.2%,特异性98.7%。
  • 关键次要终点:双向验证平衡准确率均超过88%。
  • 不良事件:原文未提及。

证据类型、等级与局限

  • 证据类型:外部验证研究(非RCT,非系统评价)。
  • 证据等级或推荐强度:原文未提及。
  • 局限:敏感性在验证方向间存在差异,可能反映队列特征或数据集大小;原文强调临床实施前需本地评估。

对床旁实践的改变

  • 该模型在外部多中心验证中保持高判别性能,提示 insult 前生理模式跨队列共享,但敏感性方向差异(93.8% vs 81.2%)表明不能直接外推至所有中心。
  • 原文结论支持在临床实施前进行本地评估,未推荐直接用于临床决策。
  • 适用人群:TBI ICU患者,具备分钟级ICP和有创动脉压监测数据。
  • 不适用人群:原文未明确排除人群,但模型训练和验证均限于TBI患者。

指南或共识出处

  • 原文未提及

以上提炼依据论文摘要(全文未开放获取,或许可不允许翻译转载);全文请点「阅读原文」。

英文原文(摘要)· 点击展开

BACKGROUND: Raised intracranial pressure is one of the most severe secondary insults after traumatic brain injury. Recognising physiological patterns before they progress to dangerous intracranial pressure elevations could enable preventive rather than reactive interventions. In our previous Uppsala single-centre study, an XGBoost machine learning model showed high discriminative performance for prediction of imminent intracranial pressure insults. The model identifies short-term physiological trajectories associated with imminent threshold crossing rather than predicting independent future clinical events. Whether these results are generalisable across centres is unknown. The aim was therefore to evaluate whether the XGBoost model trained on an Uppsala dataset maintains similar performance when applied to a multicentre TBI cohort - The CENTER-TBI High-Resolution ICU cohort.

METHODS: The XGBoost model trained on the Uppsala high frequency traumatic brain injury cohort was externally validated using data from the multicentre CENTER-TBI cohort. Minute-level averages of intracranial pressure and invasive arterial blood pressure were used to predict the onset of an intracranial pressure insult within 30 min on the basis of a 4h historical window. A bidirectional validation design was applied and performance was assessed using multiple metrics.

RESULTS: The cohorts comprised 602 patients from Uppsala and 255 patients from the CENTER-TBI database. When trained on the Uppsala cohort and tested on CENTER-TBI, the model achieved an area under the receiver operating characteristic curve (AUROC) of 97.4%, sensitivity of 93.8%, and specificity of 98.6%. Training on CENTER-TBI and testing on the Uppsala dataset yielded an AUROC of 96.1%, sensitivity of 81.2%, and specificity of 98.7%. Balanced accuracy exceeded 88% in both validation directions.

CONCLUSIONS: The XGBoost model retained high performance under external multicentre validation, indicating that pre-insult physiological patterns preceding increased intracranial pressure are shared across cohorts. Differences in sensitivity between validation directions were observed and may reflect cohort characteristics or dataset size. These findings support the need for local evaluation before clinical implementation.

来源:Neurocrit Care,查看原文。版权归原作者所有。

参考来源

  1. Galos P, Hult L, Citerio G, et al. Artificial Intelligence for Early Prediction of Intracranial Hypertension: From Single-Centre Development to Multicentre Validation. Neurocrit Care. 2026. PMID: 42873412. doi:10.1007/s12028-026-02633-w

供执业医师抢救与诊疗参考,不替代临床判断;剂量请结合体重、肝肾功能和本院药品规格复核,以指南原文为准。