Skip to content

A set of tools for extracting formattable data from clinical notes stored in electronic health record systems.

License

Notifications You must be signed in to change notification settings

uclh-criu/miade

Repository files navigation

Build Status

A set of tools for extracting formattable data from clinical notes stored in electronic health record systems. For the reference server implementation, see: miade-server.

Built with Cogstack's MedCAT package.

Contents

  1. Contributors
  2. Installing
  3. Testing
  4. Contributing
  5. Licence

Contributors

Name Email
James Brandreth j.brandreth@ucl.ac.uk
Jennifer Jiang jennifer.jiang.13@ucl.ac.uk

Installing

As the drug dosage extraction module uses Med7, you will need to download the model:

pip install https://huggingface.co/kormilitzin/en_core_med7_lg/resolve/main/en_core_med7_lg-any-py3-none-any.whl

Then install MiADE:

pip install -e .

The -e flag sets the install to auto-update, useful when developing. Remove for production.

Testing

This project uses pytest tests, which can be run with:

pytest ./tests/*

Remember, if using a virtual environment, to install pytest within your environment itself, otherwise you will be using the system python install of pytest, which will use system python and will not find your modules.

Contributing

See contributing

Licence