基于Schema的LLM抽取用于FHIR患者数字孪生
Schema-Grounded LLM Extraction for FHIR Patient Digital Twins
AI总结：
提出SG-LLM方法，通过检索增强、JSON Schema约束和验证器修复循环，从非结构化EHR中生成有效的FHIR Bundle，并在临床效用实验中优于基线。
摘要：
我们重新审视从非结构化电子健康记录（EHR）构建可互操作患者数字孪生的问题，并认为该任务更适合被视作有效FHIR Bundle的受控生成，而非抽取模块的级联。我们引入SG-LLM，一种基于schema的LLM抽取器，它(i)通过SapBERT索引检索的候选SNOMED-CT、RxNorm和LOINC代码增强提示，(ii)在直接源自FHIR R4 StructureDefinitions的JSON Schema下解码，(iii)关闭一个验证器在环修复阶段，其诊断结果作为结构化错误消息反馈。我们认为，孪生的有用性（而不仅仅是跨度级F1）才是正确的评估对象，并通过一项临床效用实验将其操作化，该实验测量了基于SG-LLM生成的FHIR Bundle与专家策划的Bundle训练的分类器在30天再入院AUROC上的差距。在MIMIC-IV和n2c2 2018 Track 2基准测试上，SG-LLM匹配或超过了强大的联合抽取和普通LLM基线，同时生成了更有效的Bundle。消融实验分离了检索、schema约束和修复循环的贡献。所有代码、提示和schema均已发布。
英文摘要：
We revisit the problem of constructing interoperable patient digital twins from unstructured electronic health records (EHRs) and argue that the task is better cast not as a cascade of extraction modules but as constrained generation of a valid FHIR bundle. We introduce SG-LLM, a schema-grounded LLM extractor that (i) augments the prompt with candidate SNOMED-CT, RxNorm, and LOINC codes retrieved through a SapBERT index, (ii) decodes under a JSON Schema derived directly from FHIR R4 StructureDefinitions, and (iii) closes a validator-in-the-loop repair stage whose diagnostics are fed back as structured error messages. We argue that the twin's usefulness, not only span-level F1, is the right object of evaluation, and operationalize this with a clinical-utility experiment that measures the gap in 30-day readmission AUROC between classifiers trained on SG-LLM-generated FHIR bundles versus expert-curated ones. On MIMIC-IV and n2c2 2018 Track 2 benchmarks, SG-LLM matches or exceeds strong joint-extraction and vanilla-LLM baselines while producing substantially more valid bundles. Ablations isolate the contributions of retrieval, schema constraint, and the repair loop. All code, prompts, and schemas are released.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
