用Transformer融合血糖与化验数据，揭示1型糖尿病的代谢亚型。
Transformer-Based Multi-Modal Temporal Embeddings for Explainable Metabolic Phenotyping in Type 1 Diabetes
- 用Transformer建模多模态时序数据，学习个体代谢状态嵌入。
- 发现5种代谢亚型，从稳定到高心血管风险，各有独特生化特征。
- 通过注意力与SHAP分析可解释，适合临床风险分层研究者参考。
1型糖尿病（T1D）代谢异质性强，传统生物标志物如糖化血红蛋白（HbA1c）难以充分表征。本研究提出一种可解释的深度学习框架，整合连续血糖监测（CGM）数据与实验室指标，学习个体代谢状态的多模态时序嵌入。通过Transformer编码器建模跨模态时序依赖，利用高斯混合模型识别潜在代谢表型。模型可解释性通过Transformer注意力可视化和基于SHAP的特征归因实现。在577名T1D患者中识别出5种潜在国内代谢表型，涵盖从代谢稳定到心血管代谢风险升高。这些表型具有显著不同的生化特征，包括血糖控制、脂质代谢、肾功能标志物及促甲状腺激素（TSH）水平差异。注意力分析显示血糖变异性为关键时序因素，SHAP分析指出HbA1c、甘油三酯、胆固醇、肌酐和TSH是表型区分的关键贡献因子。表型归属与高血压、心肌梗死、心力衰竭存在统计学上显著但较弱的关联。总体而言，该可解释的多模态时序嵌入框架揭示了生理上一致的T1D代谢亚群，支持超越单一生物标志物的风险分层。
原文摘要 · Abstract (English)
Type 1 diabetes (T1D) is a highly metabolically heterogeneous disease that cannot be adequately characterized by conventional biomarkers such as glycated hemoglobin (HbA1c). This study proposes an explainable deep learning framework that integrates continuous glucose monitoring (CGM) data with laboratory profiles to learn multimodal temporal embeddings of individual metabolic status. Temporal dependencies across modalities are modeled using a transformer encoder, while latent metabolic phenotypes are identified via Gaussian mixture modeling. Model interpretability is achieved through transformer attention visualization and SHAP-based feature attribution. Five latent metabolic phenotypes, ranging from metabolic stability to elevated cardiometabolic risk, were identified among 577 individuals with T1D. These phenotypes exhibit distinct biochemical profiles, including differences in glycemic control, lipid metabolism, renal markers, and thyrotropin (TSH) levels. Attention analysis highlights glucose variability as a dominant temporal factor, while SHAP analysis identifies HbA1c, triglycerides, cholesterol, creatinine, and TSH as key contributors to phenotype differentiation. Phenotype membership shows statistically significant, albeit modest, associations with hypertension, myocardial infarction, and heart failure. Overall, this explainable multimodal temporal embedding framework reveals physiologically coherent metabolic subgroups in T1D and supports risk stratification beyond single biomarkers.
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。