通过逐药生成与精准对齐，提升多重疾病用药推荐的准确与安全
Fine-grained List-wise Alignment for Generative Medication Recommendation
- 将用药推荐建模为逐药增删的序列决策过程
- 在多个基准数据集上达到当前最佳性能，兼顾安全与准确
- 适合需要高精度、可解释性用药推荐的临床AI研发者
准确且安全的药物推荐对有效临床决策至关重要，尤其在多重疾病患者中。现有系统依赖点级预测范式，忽视药物间的协同效应和潜在不良药物-药物相互作用（DDIs）。我们提出FLAME，一种面向大语言模型（LLMs）的细粒度列表级对齐框架，支持逐药生成药物清单。FLAME将推荐建模为序列决策过程，每一步增减单一药物。为提供细粒度学习信号，设计了基于潜在奖励塑造的步进式组相对策略优化（GRPO），显式建模DDIs并优化每种药物对整体处方的贡献。此外，通过整合结构化临床知识与协作信息，增强患者表征。在基准数据集上的实验表明，FLAME实现领先性能，兼具更高的准确率、可控的安全-准确权衡以及跨多样临床场景的强泛化能力。代码已开源：https://github.com/cxfann/Flame。
原文摘要 · Abstract (English)
Accurate and safe medication recommendations are critical for effective clinical decision-making, especially in multimorbidity cases. However, existing systems rely on point-wise prediction paradigms that overlook synergistic drug effects and potential adverse drug-drug interactions (DDIs). We propose FLAME, a fine-grained list-wise alignment framework for large language models (LLMs), enabling drug-by-drug generation of drug lists. FLAME formulates recommendation as a sequential decision process, where each step adds or removes a single drug. To provide fine-grained learning signals, we devise step-wise Group Relative Policy Optimization (GRPO) with potential-based reward shaping, which explicitly models DDIs and optimizes the contribution of each drug to the overall prescription. Furthermore, FLAME enhances patient modeling by integrating structured clinical knowledge and collaborative information into the representation space of LLMs. Experiments on benchmark datasets demonstrate that FLAME achieves state-of-the-art performance, delivering superior accuracy, controllable safety-accuracy trade-offs, and strong generalization across diverse clinical scenarios. Our code is available at https://github.com/cxfann/Flame.
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。