🩺 X-RayGuard: Pulmonary Disease Detection System

X-RayGuard is an AI-powered system for detecting lung diseases from chest X-ray images. It classifies images into three categories: COVID-19, Viral Pneumonia, and Normal. Built with TensorFlow and Gradio, this project provides a comprehensive pipeline for medical image analysis, from preprocessing to explainable predictions.

---

✨ Key Features

• Disease Classification: Detects COVID-19, Viral Pneumonia, and Normal cases.
• Explainable AI: Integrated Grad-CAM visualization to highlight decision-critical regions.
• Interactive Web Interface: User-friendly Gradio app for real-time predictions.
• Transfer Learning: Uses MobileNetV2 for efficient feature extraction.
• Detailed Metrics: Confusion matrices, classification reports, and training history plots.

---

🚀 Quick Start

Prerequisites

• Python 3.8+
• TensorFlow 2.x
• OpenCV, NumPy, Gradio

Installation

1. Clone the repository:

   git clone https://github.com/RaitonRed/X-RayGuard.git  
   cd X-RayGuard  

2. Installing Dependency's

   pip install -r requirements.txt

3. Run the web app

   cd src
   python run.py

---

📂 Project Structure

X-RayGuard/
├── .env/
├── data/
├── models/
├── notebooks/
├── results/
├── src/
│    ├── interface/
│    │   ├── __init__.py
│    │   ├── app.py
│    │   └── functions.py 
│    ├── __init__.py
│    ├── data_preprocessing.py
│    ├── evaluate.py
│    ├── grad_cam.py
│    ├── options.py
│    ├── predict.py
│    ├── run.py
│    └── train.py
├── .gitignore
├── LICENSE
├── README.md
├── requirements.txt
└── research_requirements.txt

---

🔍 Dataset

This project uses the COVID-19 Radiography Dataset. Organize the dataset as follows:

data/
├── COVID
├── NORMAL
└── VIRAL PNEUMONIA

---

📊 Model Performance

• Accuracy: 93.7% (on test data)
• Confusion Matrix:
• Training Curves:

---

🛠️ Usage Examples

1. Train the model:

python train.py

2. Generate Grad-CAM heatmaps:

python grad_cam.py --image path/to/image.png --save output.png  

3. Evaluate the model:

python evaluate.py

---

🤝 Contributing

Contributions are welcome!

• Report bugs via GitHub Issues.
• Suggest improvements or open a Pull Request.
• Improve documentation or add new features.

---

📜 License

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

---

Made with ❤️ by Raiton.

🔗 GitHub Repository | 💬 Ask a Question