LLM-Powered Personalized Glycemic Assessment in Type 2 Diabetes with Wearable Sensor Data

📄 arXiv: 2606.12699 📥 PDF

作者: Yifan Gao, Yanmin Gong, Yun Shi, Yuanxiong Guo

分类: cs.LG, cs.AI

发布日期: 2026-06-12


💡 一句话要点

提出GlyLLM框架以提升2型糖尿病个性化血糖评估

🎯 匹配领域: 支柱九:具身大模型 (Embodied Foundation Models)

关键词: 个性化医疗 糖尿病管理 可穿戴传感器 大语言模型 机器学习 血糖预测 健康数据分析

📋 核心要点

  1. 现有方法多依赖历史血糖数据,缺乏个性化信息,导致在不同人群中的表现不佳。
  2. 本文提出GlyLLM框架,利用大语言模型整合可穿戴传感器数据与个体化元数据,提升血糖动态建模能力。
  3. 实验结果显示,GlyLLM在血糖预测和糖尿病分类任务中,分别比传统方法提高了13.66%和13.08%的性能。

📝 摘要(中文)

2型糖尿病(T2D)正日益成为全球健康威胁,迫切需要有效的血糖评估以支持个性化的糖尿病护理。可穿戴传感器如连续血糖监测仪(CGM)和健身追踪器提供了许多有价值的血糖评估信息。然而,现有方法多基于传统机器学习,主要依赖历史血糖测量,忽视个体化信息,限制了在不同糖尿病人群中的表现。本文提出了GlyLLM,一个基于大语言模型(LLM)的框架,通过整合可穿戴传感器数据和结构化元数据来建模CGM基础上的血糖动态。实验结果表明,GlyLLM在血糖预测和糖尿病分类任务中均显著优于传统机器学习方法。

🔬 方法详解

问题定义:本文旨在解决现有糖尿病血糖评估方法中对个体化信息的忽视,导致模型在不同患者群体中的适用性不足。

核心思路:GlyLLM框架通过结合可穿戴传感器数据与结构化元数据,利用大语言模型的强大能力来实现个性化的血糖动态建模。

技术框架:GlyLLM的整体架构包括数据预处理模块、特征提取模块和基于LLM的决策模块,能够有效整合多种数据源。

关键创新:GlyLLM的核心创新在于利用预训练的大语言模型进行传感器数据与文本语义的抽象,显著提升了个性化评估的准确性。

关键设计:模型采用了特定的损失函数以优化血糖预测精度,并在网络结构上进行了调整,以适应多模态数据的输入。具体参数设置和网络层次结构在实验中经过反复验证。

🖼️ 关键图片

fig_0
fig_1

📊 实验亮点

实验结果表明,GlyLLM在血糖预测任务中相较于传统机器学习方法,RMSE降低了13.66%;在糖尿病分类任务中,AUROC提升了13.08%。这些结果展示了GlyLLM在个性化血糖评估中的显著优势。

🎯 应用场景

该研究在糖尿病管理领域具有广泛的应用潜力,能够为患者提供个性化的血糖监测与评估方案,进而改善糖尿病护理效果。未来,GlyLLM框架还可扩展至其他慢性疾病的个性化健康管理中,推动智能医疗的发展。

📄 摘要(原文)

Type 2 Diabetes (T2D) poses an increasing global health threat, demanding effective glycemic assessment to support personalized and improved diabetes care. Wearable sensors such as continuous glucose monitors (CGM) and fitness trackers offer many valuable insights for glycemic assessment. However, effectively analyzing these data requires integration with essential individual-level context. Existing methods are often based on traditional machine learning (ML) and rely primarily on historical blood glucose measurements and overlook personalized information, which limits their performance across diverse diabetes populations. Recent advances in large language models (LLMs) have demonstrated their ability to integrate diverse data modalities while modeling sequential dependencies, motivating the exploration of their potential for personalized glycemic assessment.In this paper, we propose GlyLLM, an LLM-powered framework for modeling CGM-based glycemic dynamics through the integration of wearable sensor data and structured metadata. GlyLLM can leverage the extensive prior knowledge of pre-trained LLMs and achieve sensor-text semantic abstraction at decision time. Experiments on two related tasks on the AI-READI dataset demonstrate that our model outperforms traditional ML methods by an average of 13.66\% in Root Mean Squared Error (RMSE) for glucose forecasting and 13.08\% in Area Under the Receiver Operating Characteristic (AUROC) for diabetes categorization. Additionally, our ablation study shows that diabetes surveys and biometric tests are more critical than other health information for glycemic assessment. Our work presents a promising step toward harnessing the power of LLMs to advance personalized glycemic assessment in T2D care.