Pneumonia Detection with Convolutional Neural Networks (CNN)

• This repository contains code to build, train, and evaluate a Convolutional Neural Network (CNN) model for pneumonia detection from chest X-ray images.
• With just a few lines of code, you can create a powerful deep learning model to assist in diagnosing pneumonia.
• This repo is designed for beginners who are starting to learn PyTorch or CNN.

Introduction

• Pneumonia is a common and potentially life-threatening condition that affects millions of people worldwide. Rapid and accurate diagnosis is crucial for effective treatment.
• This project aims to demonstrate how deep learning techniques, specifically CNNs implemented using PyTorch, can aid in the detection of pneumonia from medical images.

Notebook Contents

1. Data Loading: Utilize PyTorch's DataLoader to efficiently load and preprocess the chest X-ray images dataset.
2. Model Architecture: Define a CNN architecture tailored for pneumonia detection, leveraging PyTorch's neural network modules.
3. Training: Train the CNN model on the dataset to learn the features indicative of pneumonia presence.
4. Inference: Evaluate the trained model's performance on unseen data and infer pneumonia presence from new chest X-ray images.

Dataset

• The dataset used in this project consists of chest X-ray images collected from public repositories, containing both normal and pneumonia-affected images.
• You can find the dataset on Kaggle: Chest X-Ray Images (Pneumonia).

Usage

To run the notebook and reproduce the results:

1. Clone this repository:

   git clone https://github.com/mahadev0811/Chest-Xray-Pneumonia-Detection.git

2. Install torch and torchvision, visit PyTorch for installation instructions, only after installing torch and torchvision, run the following command:

   pip install -r requirements.txt

3. Open the notebook in Jupyter Notebook or Google Colab and run the cells.

Results

• After training the CNN model, you can expect to achieve an accuracy of over 90% on the test set on around 10 epochs of training.
• The model's performance may vary based on hyperparameters, dataset size, and other factors.

Contributions

Contributions to this project are welcome! If you have any ideas for improvements, bug fixes, or feature additions, feel free to open an issue or submit a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

Special thanks to Paul Mooney for providing the dataset used in this project.