Medical Image Classification

This repository contains a machine learning model for classifying medical images (X-rays, MRIs, CT scans) to predict whether a patient has pneumonia or is normal. The model was built using a convolutional neural network (CNN) and trained on a dataset of medical images.

Problem Statement

The goal of this project is to develop a model that can automatically classify medical images to assist healthcare professionals in diagnosing diseases like pneumonia. The model takes in images, processes them, and outputs a classification: either normal or pneumonia.

Dataset

The dataset consists of chest X-ray images that are labeled into two categories:

• NORMAL: Images that depict healthy lungs.
• PNEUMONIA: Images that depict pneumonia-infected lungs.

The dataset is divided into three parts:

• Training Set: Images used to train the model.
• Validation Set: Images used to validate the model's performance during training.
• Test Set: Images used to evaluate the final model performance.

System Overview

The system utilizes a Convolutional Neural Network (CNN) architecture to classify medical images. It works by training the model on a labeled dataset and then testing it on unseen data to predict if a chest X-ray belongs to a healthy patient or one with pneumonia.

Technologies Used

• Python: The primary programming language used.
• TensorFlow/Keras: Deep learning framework for building and training the model.
• Matplotlib: Library used for plotting training progress and results.
• NumPy: Used for numerical operations and data handling.
• OpenCV: Image processing tool to preprocess images before feeding them to the model.
• GitHub: Version control and hosting.

Setup and Installation

1. Clone the repository:

   git clone https://github.com/raeen-fatima/medical-image-classification.git
   cd medical-image-classification

2. Install the required dependencies:

   pip install -r requirements.txt

3.Download the dataset and place it in the appropriate folder (/data).

python train.py
This script will:
Load the dataset.
Preprocess the images.
Train the CNN model.
Save the trained model to disk for future use.
Evaluating the Model
To evaluate the model on test data:

    python evaluate.py

This will display the model's accuracy and other performance metrics.

Future Enhancements

Integration with Cloud Services: Deploy the model on a cloud platform for easy access.

Additional Classes: Expand the model to classify more types of diseases from medical images.

Real-Time Prediction: Implement real-time predictions using a webcam or mobile app.

License

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

Acknowledgements

Special thanks to the dataset providers for making the medical image dataset available.

Thanks to TensorFlow and Keras for providing easy-to-use frameworks for model development.