基于生理轨迹联合预测危重症风险，提升临床预警可靠性。
PGP-Clinical-TimeKAN: Prior-Guided Joint Probabilistic Forecasting of Clinical Trajectories
- 以多变量生理轨迹为输入，融合器官先验与患者特异性关系建模。
- 在MIMIC-IV数据上24小时历史+6小时预测，MAE仅比最优模型高0.52%。
- 支持概率输出且置信区间覆盖合理，适合需要可解释性的临床研究者。
临床恶化由多个相互关联的、部分观测的生理轨迹共同构成，而非单一诊断标签。我们提出PGP-Clinical-TimeKAN，一种以轨迹为核心的多变量生理联合概率预测框架。该方法结合了考虑缺失性的时序编码器、软器官系统先验、患者特异性关系结构、非线性Kolmogorov-Arnold消息传递机制以及低秩多元t分布输出头。在6,882名患者、54,694个时间窗口的MIMIC-IV冻结队列上评估24小时历史与6小时预测表现。在五个随机种子和十三种模型中，其归一化MAE（0.37727 ± 0.00029）位居第二低，RMSE（0.52656 ± 0.00034）最低。相比确定性TimeKAN，MAE降低0.52%。概率预测性能方面，边际负对数似然为0.66380，连续概率评分（CRPS）为0.27301。名义50%、80%、95%置信区间的实际覆盖率分别为0.533、0.831、0.958。移除关系结构导致最大性能下降。增加协方差秩能提升联合似然，但对点预测精度影响小。基于轨迹的风险评分仍弱于专用的GRU-D分类器（AUROC 0.603 vs 0.650），限制了当前的临床应用宣称。因此，联合轨迹预测提供了一个可解释的中间任务，但准确的生理预测本身不足以保证事件检测器的校准性。
原文摘要 · Abstract (English)
Clinical deterioration unfolds through coupled, partially observed trajectories, not a single diagnostic label. We introduce PGP-Clinical-TimeKAN, a trajectory-first framework for joint probabilistic forecasting of multivariate physiology. It combines missingness-aware temporal encoders, a soft organ-system prior, patient-specific relations, nonlinear Kolmogorov-Arnold messages, and a low-rank multivariate Student-t head. We evaluate 24-hour histories and six-hour forecasts on a frozen MIMIC-IV-derived cohort of 6,882 patients and 54,694 windows. Across five seeds and 13 models, PGP-Clinical-TimeKAN obtains the second-lowest normalized MAE (0.37727 +/- 0.00029) and the lowest RMSE (0.52656 +/- 0.00034). It reduces MAE by 0.52% relative to deterministic TimeKAN. For probabilistic forecasting, it reaches a marginal NLL of 0.66380 and a CRPS of 0.27301. Empirical coverage is 0.533, 0.831, and 0.958 for nominal 50%, 80%, and 95% intervals. Removing relational structure causes the largest ablation loss. Increasing covariance rank improves joint likelihood but has little effect on point accuracy. A trajectory-derived risk score remains weaker than a dedicated GRU-D classifier (AUROC 0.603 versus 0.650), which limits the present clinical claim. Joint trajectory forecasting therefore provides an inspectable intermediate task, but accurate physiology forecasts alone do not ensure a calibrated event detector.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。