The Deidentify app is a tool for redacting and replacing sensitive information in a document. It uses MedCAT, an advanced natural language processing tool, to identify and classify sensitive information, such as names, addresses, and medical terms.
- Redact sensitive information: The Deidentify app can automatically redact sensitive information from a document, replacing it with a placeholder value, such as "[REDACTED]".
- Replace sensitive information: Alternatively, the app can replace sensitive information with a different value, such as a random name or address, to maintain the structure and context of the original document.
- Add customizable rules: The app allows users to create custom rules for identifying and classifying sensitive information based on their specific needs and use cases.
- Batch processing: The app can process multiple documents at once, making it easy to redact or replace sensitive information in large datasets.
For out of the box models please contact: [email protected]
-
Step 1: Upload your model to the models dir.
-
Step 2: In the file .env change the
MODEL_NAME
parameter to the file name of your model. Save the changes you made to the.env
file.
MODEL_NAME = '<NAME OF MODEL HERE.zip>'
To build your own models please follow the tutorials outlined in MedCATtutorials
Note: This is currently under development
Start the Docker services by using docker-compose
. This will build the necessary Docker images and start the services.
docker-compose up
Once the services are up and running, you can access the Deidentify app in your web browser by navigating to http:https://localhost:8002.
- Install dependencies:
pip install -r requirements.txt
- Set up the database:
python manage.py migrate
This project uses frontend dependencies managed by npm. To install them, follow these steps:
-
Make sure you have Node.js and npm installed on your system.
-
Navigate to the project's app directory.
-
Run the following command to install the frontend dependencies:
npm install
npm run build
python manage.py runserver 8002