Skip to content

The Deidentify app uses MedCAT to automatically redact or replace sensitive information in documents. Customisable rules and batch processing make it easy to handle large datasets.

License

Notifications You must be signed in to change notification settings

CogStack/AnonCATDemo

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

86 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Deidentify app

Demo for AnonCAT. It uses MedCAT, an advanced natural language processing tool, to identify and classify sensitive information, such as names, addresses, and medical terms.

Example

Features

  • 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.

DeID Model

For out of the box models please contact: [email protected]

Models configured in .env ../models/ mounted into the container under /home/models/

MODEL_NAME = '<NAME OF MODEL HERE.zip>'

Build your own model

To build your own models please follow the tutorials outlined in MedCATtutorials

Note: This is currently under development

Starting the demo service

Start the Docker services by using docker-compose. This will build the necessary Docker images and start the services.

docker-compose up

About

The Deidentify app uses MedCAT to automatically redact or replace sensitive information in documents. Customisable rules and batch processing make it easy to handle large datasets.

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Vue 36.1%
  • Python 30.6%
  • CSS 14.6%
  • TypeScript 12.5%
  • Dockerfile 2.9%
  • JavaScript 1.2%
  • Other 2.1%