Detecting COVID-19 with Chest X-Ray using PyTorch

Repository: detecting-covid-xray-pytorch

📌 Project Overview

This project demonstrates a complete machine learning workflow for classifying chest X-ray images into three categories: Normal, Viral Pneumonia, and COVID-19. It was developed as part of a guided hands-on Coursera project using PyTorch and ResNet-18, with a focus on medical imaging and deep learning.

🚀 Key Features

• Custom Dataset and DataLoader: Created using PyTorch to manage image data effectively.
• Data Transformations: Applied standard preprocessing techniques like resizing, normalization, and augmentation.
• Model Training: Fine-tuned a pre-trained ResNet-18 model on a real-world chest X-ray dataset.
• Model Evaluation: Visualized predictions, tracked training and validation performance, and used accuracy as a metric.

🧠 Skills Demonstrated

• Image classification using Convolutional Neural Networks (CNNs)
• Deep learning with PyTorch
• Data augmentation and normalization
• Custom PyTorch Dataset and DataLoader
• Transfer learning with ResNet-18
• Working with medical imaging datasets

🧾 Technologies Used

• Python 3
• NumPy
• PyTorch
• Torchvision
• PIL (Python Imaging Library)
• Matplotlib

📁 Project Structure

├── complete_notebook.ipynb        # Main Jupyter notebook with code and output
├── complete_notebook.pdf          # PDF export for easy viewing
├── README.md                      # Project overview and documentation

📌 How to Use

1. Clone this repository:

   git clone https://github.com/yourusername/detecting-covid-xray-pytorch.git

2. Open complete_notebook.ipynb in Jupyter Notebook or VS Code.
3. Run the cells sequentially to see preprocessing, model training, and evaluation.
4. Alternatively, view the complete_notebook.pdf for a formatted read-only version.

📞 Contact

Khaled Jamal
Machine Learning Enthusiast & Senior Software Engineer
📧 [javaobjects@gmail.com]
🔗 LinkedIn Profile
🌍 Dubai, UAE

---

This project reflects my interest in applying machine learning to healthcare and my commitment to transitioning into AI-focused roles.