DeepTriage-CN：整合临床文本与生命体征用于老龄化人群急诊入院预测
DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.
摘要(中文译文)
急诊科对老年人的分诊具有挑战性，因为标准早期预警评分对非典型表现往往不敏感。我们开发了DeepTriage-CN，这是一个晚期融合框架，通过XGBoost分类器将护士记录的主诉的冻结BERT中文嵌入与结构化分诊生命体征相结合。该模型在8000例成人急诊就诊数据上训练，并在2000例独立就诊数据上进行时间验证。DeepTriage-CN的AUROC为0.865，与表格深度学习对照模型TabNet（0.867；DeLong p = 0.42）在统计学上无差异。该多模态模型显著优于传统临床评分（NEWS2，0.772；ESI，0.760）和仅生命体征基线模型（XGBoost，0.858；所有DeLong p < 0.001）。在老年患者（年龄≥65岁）中，该模型维持AUROC为0.852，而ESI为0.710，NEWS2为0.741，尽管性能仍与TabNet相当（0.835；p = 0.21）。在模拟30%信息性缺失并加入高斯噪声的条件下，DeepTriage-CN保留了其基线AUROC的95.4%，而TabNet为83.3%，仅生命体征XGBoost为82.8%。然而，在Youden最优分类阈值（0.28）下，阳性预测值为0.50，意味着每两个生成的警报中就有一个是假阳性。这些发现确定了一个特定的操作范围，在该范围内多模态方法具有情境优势——主要是在结构化生理数据退化时——同时确认在一般人群中，文本嵌入相较于优化的表格深度学习架构并未提供统计学显著的增量判别增益。在该框架能够指导急诊分诊实践之前，需要进行具有临床可操作终点的前瞻性多中心验证。
Abstract
Emergency department (ED) triage of older adults is challenging because standard early warning scores are often insensitive to atypical presentations. We developed DeepTriage-CN, a late-fusion framework combining frozen BERT-Chinese embeddings of nurse-recorded chief complaints with structured triage vital signs via an XGBoost classifier. The model was trained on 8000 adult ED visits and validated temporally on 2000 independent encounters. DeepTriage-CN achieved an AUROC of 0.865, statistically indistinguishable from the tabular deep-learning comparator TabNet (0.867; DeLong p = 0.42). The multimodal model significantly outperformed traditional clinical scores (NEWS2, 0.772; ESI, 0.760) and vitals-only baselines (XGBoost, 0.858; all DeLong p < 0.001). In older patients (age ≥ 65 years), the model maintained an AUROC of 0.852 versus 0.710 for ESI and 0.741 for NEWS2, though performance remained comparable to TabNet (0.835; p = 0.21). Under simulated 30% informative missingness with Gaussian noise, DeepTriage-CN retained 95.4% of its baseline AUROC compared with 83.3% for TabNet and 82.8% for vitals-only XGBoost. However, at the Youden-optimal classification threshold (0.28), positive predictive value was 0.50, meaning one in every two generated alerts is a false positive. These findings identify a specific operational envelope in which the multimodal approach confers situational advantage-primarily when structured physiological data are degraded-while confirming that text embeddings provide no statistically significant incremental discriminative gain over optimized tabular deep-learning architectures in the general population. Prospective multicenter validation with clinically actionable endpoints is required before this framework can inform ED triage practice.
如何引用
AMA
Wenjia Lin, Wenliang Chen, Guan Wei, Xiaolei Huang. DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.. Scientific reports. 2026; doi:10.1038/s41598-026-61270-7.
APA
Wenjia Lin, Wenliang Chen, Guan Wei, Xiaolei Huang (2026). DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.. Scientific reports. https://doi.org/10.1038/s41598-026-61270-7
GB/T 7714
Wenjia Lin, Wenliang Chen, Guan Wei, Xiaolei Huang. DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.[J]. Scientific reports, 2026 doi:10.1038/s41598-026-61270-7.
在 PubMed增强版 阅读全文
翻译全文 · AI 文献问答 · PDF 原文下载 · 引用导出