论文标题
多模式变压器：将临床注释与结构化EHR数据融合，用于可解释的院内死亡率预测
A Multimodal Transformer: Fusing Clinical Notes with Structured EHR Data for Interpretable In-Hospital Mortality Prediction
论文作者
论文摘要
使用结构化电子健康记录（EHR）的基于深度学习的临床决策支持一直是预测死亡率和疾病风险的活跃研究领域。同时，大量叙事临床注释提供了互补的信息，但通常不整合到预测模型中。在本文中，我们提供了一种新型的多模式变压器来融合临床注释和结构化EHR数据，以更好地预测住院死亡率。为了提高可解释性，我们提出了一种综合梯度（IG）方法来选择临床注释中的重要单词，并发现具有沙普利值的关键结构化EHR特征。这些重要的单词和临床特征可视化以帮助解释预测结果。我们还研究了域自适应预处理和任务自适应微调对临床BERT的重要性，该临床BERT用于学习临床注释的表示。实验表明，我们的模型优于其他方法（AUCPR：0.538，AUCROC：0.877，F1：0.490）。
Deep-learning-based clinical decision support using structured electronic health records (EHR) has been an active research area for predicting risks of mortality and diseases. Meanwhile, large amounts of narrative clinical notes provide complementary information, but are often not integrated into predictive models. In this paper, we provide a novel multimodal transformer to fuse clinical notes and structured EHR data for better prediction of in-hospital mortality. To improve interpretability, we propose an integrated gradients (IG) method to select important words in clinical notes and discover the critical structured EHR features with Shapley values. These important words and clinical features are visualized to assist with interpretation of the prediction outcomes. We also investigate the significance of domain adaptive pretraining and task adaptive fine-tuning on the Clinical BERT, which is used to learn the representations of clinical notes. Experiments demonstrated that our model outperforms other methods (AUCPR: 0.538, AUCROC: 0.877, F1:0.490).
