# Schema-Grounded LLM Extraction for FHIR Patient Digital Twins

arXiv: https://arxiv.org/abs/2601.05847 · HTML: https://mlsanity.com/paper/2601.05847/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 35 references that are in the map.
- First posted on arXiv 9 Jan 2026.

## Builds on (most cited first, 12 of 12)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [BioBERT: a pre-trained biomedical language representation model for biomedical text mining](https://mlsanity.com/paper/1901.08746/) (2019)
- [Large Language Models Encode Clinical Knowledge](https://mlsanity.com/paper/2212.13138/) (2022)
- [Publicly Available Clinical BERT Embeddings](https://mlsanity.com/paper/1904.03323/) (2019)
- [BioGPT: Generative Pre-trained Transformer for Biomedical Text Generation and Mining](https://mlsanity.com/paper/2210.10341/) (2022)
- [Scalable and accurate deep learning for electronic health records](https://mlsanity.com/paper/1801.07860/) (2018)
- [Generating Sequences by Learning to Self-Correct](https://mlsanity.com/paper/2211.00053/) (2022)
- [Self-Alignment Pretraining for Biomedical Entity Representations](https://mlsanity.com/paper/2010.11784/) (2020)
- [Large Language Models are Few-Shot Clinical Information Extractors](https://mlsanity.com/paper/2205.12689/) (2022)
- [Constrained Language Models Yield Few-Shot Semantic Parsers](https://mlsanity.com/paper/2104.08768/) (2021)
- [Span-based Joint Entity and Relation Extraction with Transformer Pre-training](https://mlsanity.com/paper/1909.07755/) (2019)
- [Biomedical Entity Representations with Synonym Marginalization](https://mlsanity.com/paper/2005.00239/) (2020)

## Built on by (most cited first, 1 of 1)

- [SafeGPT: Preventing Data Leakage and Unethical Outputs in Enterprise LLM Use](https://mlsanity.com/paper/2601.06366/) (2026)

## Abstract (arXiv)

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.