This repository contains a machine learning project for detecting pneumonia from chest X-ray images using Convolutional Neural Networks (CNN), Logistic Regression, and Support Vector Machines (SVM). The repository includes a Jupyter notebook for training the models, pre-trained model files, and a Streamlit web app for predicting pneumonia on uploaded X-ray images.
The Chest X-Ray Pneumonia Detection app is live on Streamlit Cloud! 🚀
Click the link below to upload X-ray images and predict if they indicate Pneumonia or Normal:
The goal of this project is to build a classification system for detecting pneumonia from chest X-ray images. The models in this project include:
- Convolutional Neural Network (CNN): Deep learning-based model for image classification.
- Logistic Regression: A traditional machine learning model for binary classification.
- Support Vector Machine (SVM): Another classification algorithm for binary outcomes.
The repository contains:
- A Jupyter Notebook that trains the models using the chest X-ray pneumonia dataset.
- Pre-trained model files for easy deployment.
- A Streamlit app for easy interaction and prediction on new chest X-ray images. The dataset used in this project is the Chest X-ray Pneumonia dataset available on Kaggle. It contains pneumonia and normal X-ray images.
https://www.kaggle.com/datasets/paultimothymooney/chest-xray-pneumonia
Make sure to download the dataset and place it in the appropriate folder if running the notebook locally.
├── models/
│ ├── cnn_model.h5 # Trained CNN model
│ ├── logistic_regression_model.pkl # Trained Logistic Regression model
│ ├── pca_transformer.pkl # Trained PCA model
├── app/
│ ├── app.py # Streamlit app for pneumonia detection
├── notebooks/
│ ├── pneumonia_detection.ipynb # Jupyter notebook with model training and comparisons
├── requirements.txt # Python dependencies for the projectCertainly! Below is a README.md template for your GitHub repository. It includes an introduction, instructions for setting up and running the notebook, usage of the models, and deploying the Streamlit app.
Chest X-Ray Pneumonia Detection This repository contains a machine learning project for detecting pneumonia from chest X-ray images using Convolutional Neural Networks (CNN), Logistic Regression, and Support Vector Machines (SVM). The repository includes a Jupyter notebook for training the models, pre-trained model files, and a Streamlit web app for predicting pneumonia on uploaded X-ray images.
Project Overview The goal of this project is to build a classification system for detecting pneumonia from chest X-ray images. The models in this project include:
Convolutional Neural Network (CNN): Deep learning-based model for image classification. Logistic Regression: A traditional machine learning model for binary classification. Support Vector Machine (SVM): Another classification algorithm for binary outcomes. The repository contains:
A Jupyter Notebook that trains the models using the chest X-ray pneumonia dataset. Pre-trained model files for easy deployment. A Streamlit app for easy interaction and prediction on new chest X-ray images. Dataset The dataset used in this project is the Chest X-ray Pneumonia dataset available on Kaggle. It contains pneumonia and normal X-ray images.
Link to the Dataset: Chest X-ray Pneumonia Dataset on Kaggle
Make sure to download the dataset and place it in the appropriate folder if running the notebook locally.
Repository Structure bash Copy code ├── models/ │ ├── cnn_model.h5 # Trained CNN model │ ├── logistic_regression_model.pkl # Trained Logistic Regression model │ ├── pca_transformer.pkl # Trained PCA model │ ├── svm_model.pkl # Trained SVM model ├── app/ │ ├── app.py # Streamlit app for pneumonia detection ├── notebooks/ │ ├── pneumonia_detection.ipynb # Jupyter notebook with model training ├── requirements.txt # Python dependencies for the project ├── README.md # This file
- Clone the Repository
git clone https://github.com/VanshajR/Pneumonia_Detection.git
cd Pneumonia_Detection- Install Dependencies
Create a virtual environment and install the necessary dependencies listed in
requirements.txt:
Create and activate virtual environment
python3 -m venv venv
source venv/bin/activate # For Windows use `venv\Scripts\activate`Install dependencies
pip install -r requirements.txt- Running the Jupyter Notebook
Open the
pneumonia_detection.ipynbnotebook in Jupyter or a compatible environment like Google Colab. Run all the cells to:
- Load the dataset.
- Train the models (CNN, Logistic Regression, and SVM).
- Save the trained models into the models/ directory.
- Running the Streamlit App Once the models are trained and saved, you can use the Streamlit app to upload new X-ray images and get predictions.
i) Navigate to the app/ folder:
cd appii) Run the Streamlit App:
streamlit run app.pyiii) Open your browser and go to http://localhost:8501 to interact with the app.
- Upload an Image to Predict
- Once the app is running, you can upload a chest X-ray image (in .jpg, .png, or .jpeg format).
- The app will show the prediction (Pneumonia or Normal) along with the confidence for each model.
- It will also display relevant metrics (accuracy, confusion matrix, etc.) for each model.
The models were evaluated based on the following metrics:
- Accuracy
- Precision
- Recall
- F1-Score
- ROC-AUC
The CNN model provides the best performance, followed by Logistic Regression. The SVM model was included as an additional benchmark but was not the top performer for this dataset.
This project demonstrates how to detect pneumonia from chest X-ray images using deep learning and traditional machine learning techniques. The Streamlit app allows users to easily upload X-ray images and get predictions using the trained models.
Feel free to modify the code, train the models on new datasets, or deploy the app for real-world use.
This project is licensed under the MIT License - see the LICENSE file for details.