INTERVenE：用于短期医疗事件预测的基于时间抽象区间的Transformer

中文摘要 AI 辅助

重症监护病房的电子健康记录（EHR）预测模型必须从稀疏且不规则的测量数据中学习，同时保留时间的临床意义并支持透明决策。我们提出INTERVenE，这是一系列Transformer架构，其输入为基于区间的、基于知识的时间抽象（KBTA），即从精心整理的医学本体中提取的命名临床概念（状态、趋势、事件、上下文）的标记流，而非未命名的区间索引或原始测量三元组。这种命名层是我们要求KBTA完成的任务：它使模型的每个标记归因从结构上解析为临床概念。INTERVenE提供两种互补变体：一种是自回归解码器，生成未来抽象轨迹并附带每一步风险读数（定位风险上升的时间点和触发事件）；另一种是双向编码器，用于单次联合风险与事件时间预测。在57078次MIMIC-IV入院数据上针对GRU-D、STraTS和KarmaLego进行评估，INTERVenE-Enc的支持加权AUPRC_w达到0.672，较最强的神经基线提升0.041，且具有不重叠的95%自助法置信区间，同时取得最佳AUROC_w（0.901）和住院时长MAE（44.4小时）。INTERVenE-Ar（在相同评估协议下的AUROC_w为0.854、AUPRC_w为0.587，属于更具挑战性的生成式读数）提供互补的标记级风险轨迹。输入表示消融实验证实该提升可跨结构化离散化迁移，确立基于KBTA的区间作为可解释的底层结构，使部署模型内的每个标记归因解析为有意义的临床概念。

英文摘要

Electronic Health Record (EHR) prediction models in the intensive care unit must learn from sparse and irregular measurements while preserving the clinical meaning of time and supporting transparent decision-making. We present INTERVenE, a family of Transformer architectures whose input is an interval-based, knowledge-based temporal abstraction (KBTA), a token stream of named clinical concepts (states, trends, events, contexts) drawn from a curated medical ontology, rather than an unnamed bin index or a raw measurement triplet. This naming layer is what we ask KBTA to do: it makes the model's per-token attributions resolve to clinical concepts by construction. INTERVenE offers two complementary variants: an auto-regressive decoder that generates future abstraction trajectories with a per-step risk readout (localizing \emph{when} and \emph{after which events} risk rises), and a bidirectional encoder for single-pass joint risk and time-to-event prediction. Evaluated on 57,078 MIMIC-IV admissions against GRU-D, STraTS, and KarmaLego, INTERVenE-Enc reaches a support-weighted AUPRC$_w$ of 0.672, improving by 0.041 over the strongest neural baseline with non-overlapping 95\% bootstrap CIs, while also taking the best AUROC$_w$ (0.901) and length-of-stay MAE (44.4\,h). INTERVenE-Ar (AUROC$_w$ $0.854$, AUPRC$_w$ $0.587$ under the same evaluation contract - a strictly harder generative readout) provides a complementary token-level risk trajectory. An input-representation ablation confirms the lift transfers across structured discretizations, positioning KBTA-based intervals as the interpretable substrate that makes per-token attributions resolve to meaningful clinical concepts within the deployed model.

发表机构

• The Stein Faculty of Computer and Information Science, Ben-Gurion University of the Negev（内盖夫本-古里安大学斯坦因计算机与信息科学学院）

机构由 AI 辅助整理，请以论文原文为准。