Heart Stroke Predictions Model In Production

Users : 🩺

• Medical professionals 👨‍⚕️
• Clinics / hospitals 🏥
• Medical devices 🔬

Usage Description: 🫀

After providing the necessary information to the health professionals of the user or inputting his or her personal & health information on the medical device or the Web Interface. Our model will use the the information provided by the user above to predict the probability of him having a stroke. After that the Web interface will display a detailed result about the patient status and possible precautions or advices to visit a professional

Features:

Our application will feature a :

• Web interface & Data Search Interface using Streamlit
• Prediciton API using FastApi
• Machine Learning Model as Python Package "stroke-pred-p0w11'
• Data Storage unit using PostgresSQl & Sqlalchmey
• Data Ingestion job using Airflow to collect our data based on the user inputs.
• Prediction monitoring dashboard using Gafana

Dataset:

• 11 clinical features for predicting stroke events:
  https://www.kaggle.com/datasets/fedesoriano/stroke-prediction-dataset
  

Postgres Database Setup :

1. Make sure to install database dependencies [psycopg2, python-dotenv, sqlalchemy]
   -Check stroke_heart_prediciton/requirements.txt (Remark For Mac, Linux Users psycopg2-binary) 👈
   
2. Create a (.env) file in the main Root => stroke_heart_prediciton/.env
3. (.env) File Should Contain: ❗

[POSTGRES_DB]
POSTGRES_USER=[User]
POSTGRES_PASSWORD=[Password]
POSTGRES_SERVER=[Server]
POSTGRES_PORT=[Port]
POSTGRES_DB=[Database]

[FastApi]
BACKEND_SERVER =[Server]

4. Open terminal and go to Cd stroke_heart_prediciton/postgres
5. Run Python createdb.py to create the tables & relationships in your database

Airflow ( Follow the steps in Repo ) ⏲️

Airflow Repo - README.md
Link to Airflow

Grafana ( Follow the steps in Repo ) 🌀

Grafana Repo - README.md
Link to Grafana

Heroku Streamlit

Link to Web Interface

Execute Program Locally:

1. Cd stroke_heart_prediciton/stroke_api; uvicorn main:app --host 0.0.0.0 --port 8005;

2. streamlit run web_interface.py --server.port 8010;

System Architecture: 🧱