MIMICWizard is an open-source web application that enables intuitive exploration of the MIMIC-IV critical care database by clinicians and researchers without programming expertise
Note
When using this resource, please cite the associated article : (submitted, waiting for review)
Large public EHR datasets like MIMIC-IV contain rich, longitudinal patient data but are difficult to navigate for users lacking data science training.
MIMICWizard addresses this gap by providing an interactive dashboard for patient-level data visualization, cohort query and creation, statistical summarization for MIMIC-IV database.
Originally developed for clinicians, MIMICWizard supports hypothesis generation and data exploration in a hospital context.
- Visual exploration of patient data from MIMIC-IV.
- Cohort creation and exploration for hypothesis generation
- Stratification and descriptive statistics on population and subgroups
- Advanced exploration with search engine including events, ICD and demographics
- Demo mode with a lightweight dataset.
- Open-source and community-driven.
A demo version of the application is available at https://mimicwizard-demo.lcr.datailor.eu/.
This demo version use MIMIC-IV demo database, an openly-available demo of MIMIC-IV containing a subset of 100 patients.
MIMICWizard is based on R Shiny and interact with a PostgreSQL database. Detailed setup instruction are available on the dedicated documentation page.
More information about the application and usage are available on the ReadTheDocs documentation.
Important
Access to the full MIMIC-IV dataset requires credentialing and completion of a data use agreement. Visit PhysioNet for more information.
v0.8.0 (2025-11-05) Clinical notes and extended data
General
- Support for MIMIC-IV clinical notes (discharge), activate the flag in the global.R file
Patient Explorer
- Adding hadm/stay start and end time visualization on timeline
- Add extended data when visualize hospitalization, get access to lab and microbiology result when data is consitent with hospital stay but not linked to a hadm_id (useful for ED lab measurement)
- Bugfixes
v0.7.1 (2025-06-27) Bugfixes and optimization
v0.7.0 (2025-06-20) Public release associated with online demo version
v0.6.4 (2025-03-13) First closed full version beta-testing
This application was developed by the Laboratoire Commun de Recherche - HCL-bioMérieux (Lyon, France), a joint research initiative between Edouard Herriot Hospital and bioMérieux.
📫 For questions, suggestions, or collaborations, feel free to open an issue or contact the research team via the GitHub repository.
Warning
This is not an official product of bioMérieux or Hospices Civils de Lyon. The software is provided "as is" without warranty under GPLv3 license.
We welcome contributions from the community! To contribute:
- Open a issue about your concerns, problems or feature proposal
- Fork the repository and open a Pull Request with your code suggestion