ShoQPred - AI for Shock Prediction in ICU Validated across Continents

Predicting Hemodynamic shock using physiological vitals time-series data

ShoQPred is Deployable Human Centric Software Suite for hemodynamic shock prediction which leverages AI and ML methods on physiological vitals time-series data to predict hemodynamic shock prediction upto 3 to 12 hours before the event.

Requirements

1. Python 3.6+ / Django
2. Tsfresh
3. Sklearn (extended requirements in requirements.txt)

or Docker for Docker deployment

Deployment

Docker

Readily available installation, user can pull the docker image from docker hub, here are the instruction to use it.

Create DockerHub Account and login on docker command line using docker login. Pull and run app.

docker pull raptor4/shoqpredxgb:latest
docker run --name shoqpred-app -p 5005:5005 raptor4/shoqpredxgb:latest

App will run on localhost:5005

Native Python

Native System Deployment Reqirements

System must have:-

1. Python 3.6+
2. Tsfresh

Clone the repository, install requirements and run app.

git clone https://github.com/SAFE-ICU/ShoQPred
pip3 install -r requirements.txt
python3 manage.py runserver

Sample input files are available in sample data.