Working with CogStack

This repo has been moved to the CogStack organisation. No further updates will be made to this repo

This repository contains all tools relevant to interacting with an NHS deployment of CogStack.

It contains easy to follow templates and instructions to interact and search CogStack.

NOTE These scripts have been moved to the main repo in the CogStack Org

Tasks left TODO:

Setup

Users can follow these steps to quickly setup and deploy this repository on their machine.

Any code to enter in these instructions will be represented as code to enter.

Please replace anything within <Enter information here> with your own specific details.

Step 1: Clone this repository locally

Enter the directory where you would like to store these files. cd path/to/where/you/want/this/repository
Clone the online repository: git clone https://github.com/antsh3k/working_with_cogstack.git

Further instructions and self-help with git and git clone. Please visit this link.

If you choose to use github desktop rather than the terminal please refer to the official github desktop guides.
Optional: To update to the latest release of this repository: git pull

Step 2: Creating a virtual environment and required packages

(Requires Python 3.7+)

Windows

Create a new virtual env: python3 -m venv venv
Load the virtual environment: .\venv\Scripts\activate
Install relevant packages and libraries: pip install -r requirements.txt

Linux/MAC OS

Create a new virtual env: python3 -m venv venv
Load the virtual environment: source venv/bin/activate
Install relevant packages and libraries: pip install -r requirements.txt

Optional: If no jupyter instance is installed.

In the main folder of this repository. Activate your virtual environment, using the (Step 2) command from your respective OS.
Start jupyter notebook: jupyter notebook

Step 3: Enter credentials and Login details

In the main folder of this repository you must populate the credentials.py file with your own CogStack hostnames, username and passwords.

If you have any questions or issues obtaining these details please contact your local CogStack administrator.

The credentials.py contents:

hosts = []  # This is a list of your cogstack elasticsearch instances.

# These are your login details (either via http_auth or API)
username = ''
password = ''
api_username = ''
api_password = ''

How to search using CogStack

This directory contains the basics search templates.

For further information on CogStack please visit their github or wiki page.

How to create a watcher

This directory contains the basics watcher job templates.

MedCAT

An overview of this process is shown below.

Further information about MedCAT can be found from their github or via their official documentation here.

General MedCAT tutorials can be found here.

Demo

A demo application is available at MedCAT. This was trained on MIMIC-III to annotate SNOMED-CT concepts. Note: No supervised training has been provided to this model and therefore should only be used for demonstration purposes only.

MedCAT Citation

@ARTICLE{Kraljevic2021-ln,
  title="Multi-domain clinical natural language processing with {MedCAT}: The Medical Concept Annotation Toolkit",
  author="Kraljevic, Zeljko and Searle, Thomas and Shek, Anthony and Roguski, Lukasz and Noor, Kawsar and Bean, Daniel and Mascio, Aurelie and Zhu, Leilei and Folarin, Amos A and Roberts, Angus and Bendayan, Rebecca and Richardson, Mark P and Stewart, Robert and Shah, Anoop D and Wong, Wai Keong and Ibrahim, Zina and Teo, James T and Dobson, Richard J B",
  journal="Artif. Intell. Med.",
  volume=117,
  pages="102083",
  month=jul,
  year=2021,
  issn="0933-3657",
  doi="10.1016/j.artmed.2021.102083"
}

antsh3k/working_with_cogstack