aiqbench analysis
REFINE introduces a novel framework that leverages frozen large language models and reinforcement learning to personalize medical concept representations within text-attributed knowledge graphs. By dynamically allocating graph context budgets for individual patients, the method effectively aligns semantic information with specific clinical trajectories, yielding robust improvements in electronic health record prediction tasks. This approach addresses critical limitations in existing encoders that treat patient data uniformly, offering a significant step forward in context-aware medical AI modeling.
REFINE: LLM Refinement over Budgeted Text-Attributed Graphs for Personalized Medical Concept Representation
arXiv:2609.04415v1 Announce Type: cross Abstract: Learning rich medical concept representations is essential for EHR prediction. Text-attributed knowledge graphs (TKGs) provide a natural foundation by organizing heterogeneous medical relations together with textual semantics. However, most existing encoders process concepts uniformly across patients, despite the fact that a code's meaning and predictive value depend on patient-specific clinical context and trajectory. Learning patient-personalized concept representations from TKGs introduces two key challenges: (1) deciding how much KG context to incorporate for each observed code, and (2) aligning semantic information with the patient-specific relational structure. We propose REFINE, a KG-aware budgeted LLM graph refinement framework for patient-personalized medical concept encoding. Starting from a global TKG, REFINE constructs patient-specific temporal graphs. A sequential reinforcement learning policy selects a personalized KG expansion budget for each observed code. The resulting patient graph is processed by a heterogeneous GNN to capture relation-aware structural dependencies, while a frozen LLM uses graph-aware soft prompts to semantically refine concept representations. Experiments on MIMIC-III and MIMIC-IV show that REFINE consistently improves diverse EHR backbones, outperforms strong baselines, and demonstrates robust gains across component ablation, KG selection, and data insufficiency.
Abstract:Learning rich medical concept representations is essential for EHR prediction. Text-attributed knowledge graphs (TKGs) provide a natural foundation by organizing heterogeneous medical relations together with textual semantics. However, most existing encoders process concepts uniformly across patients, despite the fact that a code's meaning and predictive value depend on patient-specific clinical context and trajectory. Learning patient-personalized concept representations from TKGs introduces two key challenges: (1) deciding how much KG context to incorporate for each observed code, and (2) aligning semantic information with the patient-specific relational structure. We propose REFINE, a KG-aware budgeted LLM graph refinement framework for patient-personalized medical concept encoding. Starting from a global TKG, REFINE constructs patient-specific temporal graphs. A sequential reinforcement learning policy selects a personalized KG expansion budget for each observed code. The resulting patient graph is processed by a heterogeneous GNN to capture relation-aware structural dependencies, while a frozen LLM uses graph-aware soft prompts to semantically refine concept representations. Experiments on MIMIC-III and MIMIC-IV show that REFINE consistently improves diverse EHR backbones, outperforms strong baselines, and demonstrates robust gains across component ablation, KG selection, and data insufficiency.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2609.04415 [cs.LG]
(or arXiv:2609.04415v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2609.04415 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 278 words
Related coverage
Performance Profiling on AMD GPUs – Part 5: Profiling-Driven Kernel Optimization with an AI Code-Assist Tool
AMD · Jul 16, 2026
[AINews] AI Engineer NYC: Gemini 4 Argon, Vals AI Benchmarks
Latent Space · Oct 14, 2026
100 things we announced at I/O 2026
Google · May 20, 2026
2026.7: Automations that speak your language
Home Assistant · Jul 1, 2026