MIMIC-Note-Extract

A Large-Scale Structured Dataset Extracted from MIMIC-IV-Note Discharge Summaries

ML4H 2026 · Proceedings Track

Kristian Schwethelm^1,* Jonas Kuntzer^1,* Conrad Ketzer² Eva Wende³ Moritz Schlager^1,4 Daniel Rückert^1,4,5 Georgios Kaissis⁶ Friederike Jungmann^1,3 Paul Hager¹

¹AI in Healthcare and Medicine, TUM and TUM University Hospital · ²Trauma Surgery, TUM and TUM University Hospital · ³Diagnostic and Interventional Radiology, TUM and TUM University Hospital · ⁴Munich Center for Machine Learning · ⁵Imperial College London · ⁶Hasso Plattner Institute, University of Potsdam

^*Equal contribution

OpenReview arXiv Dataset code Benchmark

TL;DR

We release MIMIC-Note-Extract (MNE), a structured clinical dataset extracted from 300,000+ MIMIC discharge summaries using schema-guided LLMs. We use MNE to build a clinical decision-making benchmark for LLM agents.

Read full abstract Hide abstract

Abstract

MIMIC is one of the most widely used public hospital datasets for machine learning on electronic health records, yet clinically important information, such as past medical history, allergies, and diagnostic findings, remains confined to free-text discharge summaries. Studies requiring this information must therefore build their own extraction pipelines, duplicating substantial effort across projects.

We release MIMIC-Note-Extract (MNE), a large-scale dataset that uses a hybrid extraction pipeline combining deterministic parsing with schema-guided large language model generation to transform 16 sections from 300,720 discharge summaries into structured tables. MNE adds information that has no structured MIMIC counterpart and enriches related tables with broader coverage and greater clinical detail, including 10 times more patient-history entries and medication histories for more than twice as many admissions. Validation by medical doctors confirms extraction accuracy above 92% across sections. Using MNE, we construct a clinical decision-making benchmark for LLM agents, demonstrating how reusable structured representations of clinical narratives reduce data curation effort.

See how it works

Citation

Paper BibTeX

@inproceedings{schwethelm2026mne,
  title={{MIMIC}-Note-Extract: A Large-Scale Structured Dataset Extracted from {MIMIC}-{IV}-Note Discharge Summaries},
  author={Kristian Schwethelm and Jonas Kuntzer and Conrad Ketzer and Eva Wende and Moritz Schlager and Daniel Rueckert and Georgios Kaissis and Friederike Jungmann and Paul Hager},
  booktitle={Machine Learning for Health 2026},
  year={2026},
  url={https://openreview.net/forum?id=JrkNHy2wDR}
}

Dataset BibTeX

@article{kuntzer2026mne,
  author = {Jonas Kuntzer and Kristian Schwethelm and Conrad Ketzer and Eva Wende and Moritz Schlager and Daniel Rueckert and Georgios Kaissis and Friederike Jungmann and Paul Hager},
  title = {{MIMIC-IV-Ext MIMIC Note Extract: A Large-Scale Structured Dataset Extracted from MIMIC-IV-Note Discharge Summaries}},
  journal = {{PhysioNet}},
  year = {2026},
  month = oct,
  note = {Version 1.0},
  doi = {10.13026/cq50-rd11},
  url = {https://doi.org/10.13026/cq50-rd11}
}