🧠 DR Analyzer — Diabetic Retinopathy Classification Web App

DR Analyzer is a lightweight, Flask-based web application that classifies retinal fundus images into different stages of Diabetic Retinopathy (DR). Designed for clinical insight and academic utility, it leverages a deep learning model trained on real-world datasets to provide accurate and immediate feedback.

⚕️ Empowering healthcare professionals and researchers with AI-driven diagnostics for retinal health.

---

🚀 Features

• Upload and analyze fundus images directly through the browser.
• Classifies into five stages: No DR, Mild, Moderate, Severe, and Proliferative DR.
• Real-time inference via a lightweight Flask backend.
• Simple, intuitive web interface.
• Easily extendable to support new models or additional ophthalmic conditions.

---

🧠 Model Overview

• Based on a custom DRNet architecture, optimized for retinal image analysis.

---

🐳 Docker Deployment

Prerequisites

• Docker installed on your system
• Docker Compose installed on your system

Building and Running with Docker

1. Build and start the container:

   docker-compose up -d

2. Access the application: Open your browser and go to:

   http://localhost:5000
   

3. Stop the container:

   docker-compose down

Building without Docker Compose

If you prefer not to use Docker Compose:

1. Build the Docker image:

   docker build -t dr-analyzer .

2. Run the container:

   docker run -d -p 5000:5000 --name dr-analyzer dr-analyzer

3. Stop and remove the container:

   docker stop dr-analyzer
   docker rm dr-analyzer

• Trained on large-scale datasets like APTOS and EyePACS.
• Includes preprocessing (grayscale normalization, contrast enhancement, resizing).
• Utilizes multi-scale features and attention mechanisms.
• Optional use of perceptual loss for enhanced visual consistency in image-driven diagnosis.

---

📁 Repository Structure

├── static/
├── templates/
├── model/
├── app.py
├── requirements.txt
├── README.md
├── CONTRIBUTIONS.md
├── CODE_OF_CONDUCT.md

---

⚙️ Getting Started

Run DR Analyzer locally by following the steps below:

1. Clone the Repository

git clone https://github.com/nameishyam/mini-webapp.git
cd mini-webapp

2. Create a Virtual Environment

python3 -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Required Packages

pip install -r requirements.txt

4. Add Pretrained Model

Download or place your trained model (e.g., drnet_model.pth) into the model/ folder.

⚠️ The model file is not included due to size limits. Contact the author or train your own using the APTOS/EyePACS dataset.

5. Launch the Application

python app.py

Visit the app at: http://127.0.0.1:5000

---

🧪 Sample Usage

Upload retinal fundus images in .jpg, .jpeg, or .png format. The prediction will appear immediately after upload, along with class labels.

---

📊 Roadmap

• ✅ Core classification via DRNet
• 🚧 Grad-CAM-based visual explanations
• 🚧 Deployment via Docker and CI/CD
• 🚧 REST API version for mobile and external use

---

🙌 How to Contribute

We welcome and value contributions from the community. Whether it's improving the UI, enhancing the model, fixing bugs, or writing documentation—your support matters.

• Review the 📘 CONTRIBUTIONS.md for guidelines.
• Please follow our 🤝 CODE_OF_CONDUCT.md to maintain a respectful and inclusive environment.

---

📄 License

This project is licensed under the MIT License. See LICENSE for details.

---

📬 Contact

• GitHub: @nameishyam
• Email: geddamgowtham4@gmail.com