Diabetic Retinopathy Detection

Introduction

This repository contains the AI Ambassadors Project of Microsoft Learn Student Ambassadors and Hacksagon @ IIITM Gwalior. The project is about Diabetic Retinopathy Image Detection Model. The project is divided into two parts:

• An EfficientB5 Model trained on the APTOS 2019 dataset. Achieved a Cohen Kappa Score of 0.96 on the complete dataset. DR Detection Model | Huggingface

• A Web application, powered by Django to detect diabetic retinopathy from retinal fundus images using the trained ResNet50 model on HuggingFace. Hosted on Azure Web Apps.

Presentation

The presentation for the project can be found at: Oculus SA - Diabetic Retinopathy Detection. (Access restricted to Microsoft Learn Student Ambassadors)

Usage

1. Clone the repository and change directory :

   git clone https://github.com/Polymath-Saksh/DR_Detection.git
   cd DR_Detection

2. Set up environment variables:

   Create a .env file in the root directory and add the following variables.

   # then edit .env with:
   AZURE_ACCOUNT_KEY=<your_account_key>
   AZURE_ACCOUNT_NAME=<your_account_name>
   AZURE_CONTAINER=<your_container_name>
   AZURE_STORAGE_CONNECTION_STRING=<your_connection_string>
   AZURE_URL=<your_url>
   DEBUG=<your_debug_value>
   SECRET_KEY=<your_secret_key>
   USE_AZURE_STORAGE=<your_use_azure_storage_value>

3. Create a virtual environment and install dependencies:

   python -m venv venv
   venv\Scripts\Activate
   pip install -r requirements.txt

4. Apply migrations and run locally:

    python manage.py migrate
    python manage.py collectstatic
    python manage.py runserver

For production deployment, configure Azure Web Apps or your preferred host.

Contributors (Team Oculus)

• Saksham Kumar
• Rhythm Narang
• Hitesh Khatwani
• Vaishnavi Ahire
• Aloukik Joshi

Acknowledgments

• EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks by Mingxing Tan, Quoc V. Le, 2019.

License

This project is licensed under the Creative Commons Attribution-NonCommercial 4.0 International License (CC BY-NC 4.0), which permits others to share and adapt the material for non-commercial purposes, provided that appropriate credit is given to the original author.