Git Product home page Git Product logo

mimic-clinical-decision-making-dataset's Introduction

MIMIC Clinical Decision Making

🔥 New Addition: Llama 3 has been added to the leaderboard! 🔥

This repository contains the code to create the MIMIC Clinical Decision Making (MIMIC-CDM) Dataset.

Visit https://huggingface.co/spaces/MIMIC-CDM/leaderboard to check out the current leaderboard. I will update this as new models are released. If you would like a model to be tested and put on the board, please write me an email at paul (dot) hager (at) tum (dot) de.

Environment

To setup the environment, create a new virtual environment of your choosing with python=3.10 and then run:

pip install --no-deps -r requirements.txt

Option 1: Generate the dataset from MIMIC-IV-Ext-CDM directly

The most straightforward way of using the data is by downloading it from https://physionet.org/content/mimic-iv-ext-cdm/ and then converting it to the necessary format. To convert it, first set the base_new variable at the top of ConvertPhysionet.py to the folder containing the downloaded MIMIC-IV-Ext-CDM dataset. Then run:

python ConvertPhysionet.py

Option 2: Generate the dataset from scratch from MIMIC-IV v2.2

The more laborious and time-intensive way of creating the dataset is from the raw MIMIC-IV v2.2 download. First set the following two path variables in CreateDataset.py:

  1. base_mimic should be set to the parent folder of your MIMIC-IV download. This folder should contain the hosp and note folders.
  2. base_new should be set to a folder where the generated files will be saved.

Then run:

python CreateDataset.py

Citation

If you found this code and dataset useful, please cite our paper and dataset with:

Hager, P., Jungmann, F., Holland, R. et al. Evaluation and mitigation of the limitations of large language models in clinical decision-making. Nat Med (2024). https://doi.org/10.1038/s41591-024-03097-1

@article{hager_evaluation_2024,
	title = {Evaluation and mitigation of the limitations of large language models in clinical decision-making},
	issn = {1546-170X},
	url = {https://doi.org/10.1038/s41591-024-03097-1},
	doi = {10.1038/s41591-024-03097-1},,
	journaltitle = {Nature Medicine},
	shortjournal = {Nature Medicine},
	author = {Hager, Paul and Jungmann, Friederike and Holland, Robbie and Bhagat, Kunal and Hubrecht, Inga and Knauer, Manuel and Vielhauer, Jakob and Makowski, Marcus and Braren, Rickmer and Kaissis, Georgios and Rueckert, Daniel},
	date = {2024-07-04},
}

Hager, P., Jungmann, F., & Rueckert, D. (2024). MIMIC-IV-Ext Clinical Decision Making: A MIMIC-IV Derived Dataset for Evaluation of Large Language Models on the Task of Clinical Decision Making for Abdominal Pathologies (version 1.0). PhysioNet. https://doi.org/10.13026/2pfq-5b68.

@misc{hager_mimic-iv-ext_nodate,
	title = {{MIMIC}-{IV}-Ext Clinical Decision Making: A {MIMIC}-{IV} Derived Dataset for Evaluation of Large Language Models on the Task of Clinical Decision Making for Abdominal Pathologies},
	url = {https://physionet.org/content/mimic-iv-ext-cdm/1.0/},
	shorttitle = {{MIMIC}-{IV}-Ext Clinical Decision Making},
	publisher = {{PhysioNet}},
	author = {Hager, Paul and Jungmann, Friederike and Rueckert, Daniel},
	urldate = {2024-07-04},
	doi = {10.13026/2PFQ-5B68},
	note = {Version Number: 1.0
Type: dataset},
}

mimic-clinical-decision-making-dataset's People

Contributors

paulhager avatar

Stargazers

 avatar Shivshankar Umashankar avatar  avatar Amrin Kareem avatar Mert Sayar avatar David avatar Sami Nas avatar Jieli Zhou avatar  avatar Junwei Yang avatar Ismaël RAYMOND avatar Yun Li avatar

Watchers

 avatar

mimic-clinical-decision-making-dataset's Issues

'patho_ids' is empty in ConvertPhysionet.py

Thank you very much for sharing such a meaningful dataset. However, it seems that there are some bugs in ConvertPhysionet.py that need to be fixed. In ConvertPhysionet.py, from Line 130 to Line 152, 'patho_ids' is initially set to four empty lists, but then in Line 142:

for _id in patho_ids["appendicitis"]:

the enumeration of patho_ids begins, which directly results in ConvertPhysionet.py generating four empty files.

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.