arXiv cs.CL· Bayram Ayadi, Florian Geissler, Johanna Schmidhuber·· 13 小时前AI 评分29
从结构化记录到患者叙事：LLM 生成健康摘要如何保持临床对齐
From Structured Records to Patient Narratives: Preserving Clinical Alignment in LLM-Generated Health Summaries
AI 导读
研究提出一种结合结构化证据条件与 RAG 的约束生成框架，用于月经健康监测中的证据支撑型健康摘要。在三个 LLM 生成的 4,500 条摘要上，显式 grounding 约束持续提升临床对齐，尤其改善不确定性保留，并在对照检索证据评估忠实度时总体保持或提升忠实度，但遗漏情况因模型而异。临床对齐的提升以阅读难度和医学术语增加为代价。
正文 · 原文
Abstract:As wearable and digital health systems increasingly generate longitudinal physiological records, communicating these observations in patient-accessible language remains a significant challenge. Large Language Models (LLMs) offer a promising mechanism for translating structured clinical information into readable narratives, but this process may introduce semantic drift, where clinically precise meaning is distorted during simplification. We investigate evidence-grounded health summarization for menstrual health monitoring using a constrained generation framework that combines structured evidence conditioning with Retrieval-Augmented Generation (RAG) using a prompt-based constrained generation framework. We evaluate semantic alignment using Natural Language Inference (NLI), atomic factuality metrics, and readability measures. Across three LLMs and 4,500 generated summaries, explicit grounding constraints consistently improve clinical alignment, particularly by improving uncertainty preservation and, once faithfulness is evaluated against retrieved evidence, generally preserving or improving faithfulness, while effects on omission remain model-dependent. We also find that improved clinical alignment comes at the cost of increased reading difficulty and medical terminology. These findings highlight the need to jointly evaluate faithfulness, uncertainty preservation, and readability when developing patient-facing medical LLM systems.
| Comments: | 18 pages, 3 figures |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.31467 [cs.CL] |
| (or arXiv:2610.31467v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.31467 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Bayram Ayadi [view email]
[v1]
Thu, 22 Oct 2026 15:58:57 UTC (1,697 KB)
来源：arXiv cs.CL · arxiv.org