PGP-Clinical-TimeKAN：先验引导的临床轨迹联合概率预测
PGP-Clinical-TimeKAN: Prior-Guided Joint Probabilistic Forecasting of Clinical Trajectories
浏览论文内容
中文总结 AI 辅助
提出PGP-Clinical-TimeKAN框架，结合缺失感知编码器、器官先验和低秩Student-t头，实现多变量生理轨迹的联合概率预测，在MIMIC-IV队列上取得最优RMSE，但轨迹风险评分弱于专用分类器。
中文摘要 AI 辅助
临床恶化是通过耦合的、部分观测的轨迹展开的，而非单一的诊断标签。我们提出了PGP-Clinical-TimeKAN，一个轨迹优先的框架，用于多变量生理学的联合概率预测。它结合了缺失感知的时间编码器、软器官系统先验、患者特定关系、非线性Kolmogorov-Arnold消息以及低秩多元Student-t头。我们在一个冻结的MIMIC-IV衍生队列上评估了24小时历史和6小时预测，该队列包含6,882名患者和54,694个窗口。在五个种子和13个模型上，PGP-Clinical-TimeKAN获得了第二低的归一化MAE（0.37727 ± 0.00029）和最低的RMSE（0.52656 ± 0.00034）。相对于确定性TimeKAN，它将MAE降低了0.52%。对于概率预测，它达到了边际NLL 0.66380和CRPS 0.27301。在名义50%、80%和95%区间下，经验覆盖率分别为0.533、0.831和0.958。移除关系结构导致最大的消融损失。增加协方差秩改善了联合似然，但对点精度影响不大。轨迹衍生的风险评分仍然弱于专门的GRU-D分类器（AUROC 0.603对比0.650），这限制了当前的临床主张。因此，联合轨迹预测提供了一个可检查的中间任务，但仅凭准确的生理预测并不能确保校准的事件检测器。
英文摘要
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.
发表机构
- Tianjin University（天津大学）
机构由 AI 辅助整理，请以论文原文为准。
