发表机构
MIMIC-IV; eICU-CRD（MIMIC-IV; eICU-CRD）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
研究针对ICU时间序列预测问题，提出DRIFT混合框架，直接模型主预测，递归行动条件模型校正。在MIMIC-IV和eICU-CRD数据上评估，结果显示其在降低MAP平均绝对误差上有优势，在特定审核及稳健性实验中表现良好。
AI 中文摘要
许多时间序列预测不仅依赖先前观测，还取决于预测期内指定的行动。在重症监护病房（ICU）中，未来生命体征和实验室值受血管加压药等治疗影响。但一次性预测完整未来序列的模型很少利用这些治疗，自回归模型会累积误差。我们引入DRIFT，一种混合框架，直接模型产生主要预测，递归的行动条件模型进行约束校正。我们在MIMIC-IV的6046例入院病例和eICU-CRD的8345例入院病例上评估DRIFT。在8小时、24小时和48小时预测终点上，相对于行动条件时间融合变压器（TFT-action），DRIFT使MIMIC-IV上平均动脉压（MAP）的平均绝对误差降低0.673%，在eICU-CRD上的比较模型中实现了最低相应误差。在MIMIC-IV仅限于供应治疗序列改变窗口的审核中，DRIFT在8小时和24小时的观察目标MAP误差低于TFT-action。治疗序列改变使DRIFT的MAP误差比TFT-action增加得更多，预测变化主要发生在供应路径分歧后。在单独的稳健性实验中，在强调总体终点误差、MAP误差或两者同等的三个共享检查点选择规则下，MAP优势持续存在。
英文摘要
Many time-series forecasts depend not only on prior observations but also on actions specified during the forecast period. In intensive care units (ICUs), future vital signs and laboratory values are influenced by treatments such as vasopressors. However, models that predict the full future sequence all at once make little use of these treatments, whereas autoregressive models can accumulate errors. We introduce DRIFT, a hybrid framework in which a direct model produces the primary forecast and a recursive, action-conditioned model contributes constrained corrections. We evaluate DRIFT on 6,046 admissions from MIMIC-IV and 8,345 admissions from eICU-CRD. Averaged across the 8-, 24-, and 48-hour forecast endpoints, DRIFT reduces mean absolute error for mean arterial pressure (MAP) by 0.673% relative to an action-conditioned Temporal Fusion Transformer (TFT-action) on MIMIC-IV and achieves the lowest corresponding error among the compared models on eICU-CRD. Although the overall accuracy improvement is modest, a MIMIC-IV audit restricted to windows in which the supplied treatment sequence was altered showed that DRIFT achieved lower observed-target MAP error than TFT-action at 8 and 24 hours. Treatment-sequence alteration increased DRIFT's MAP error by 0.21-0.26 mmHg more than it increased TFT-action's error, with prediction changes occurring primarily after the supplied paths diverged. In a separate robustness experiment, the MAP advantage persisted under three shared checkpoint-selection rules emphasizing overall endpoint error, MAP error, or both equally.
Comments34 pages, 1 figure; extended technical appendices included
