Brain Tumor Classification

This repository contains a deep learning-based solution for classifying brain tumors using MRI images. The model is trained to classify images into four categories: No Tumor, Pituitary, Glioma, Meningioma

Table of Contents

• Brain Tumor Classification
  • Table of Contents
  • Overview
  • Dataset
  • Requirements
  • Training
  • Evaluation
  • Streamlit App
    • Functionality:
  • Project Demo
  • Usage
  • Results
  • Acknowledgments
  • Kaggle Notebook

Overview

This project uses a Convolutional Neural Network (CNN) implemented in PyTorch to classify brain MRI images. The model architecture consists of multiple convolutional, batch normalization, max-pooling layers followed by fully connected layers.

Dataset

The dataset used is the Brain Tumor MRI Dataset available on Kaggle. It contains MRI images for training and testing the model.

• Brain Tumor MRI Dataset

Requirements

• Python 3.x
• PyTorch
• Torchvision
• NumPy
• Scikit-learn
• Matplotlib
• Seaborn
• Streamlit

Training

The training script preprocesses the images, defines the model architecture, and trains the model.

1. Preprocessing: Images are resized and normalized.
2. Model Architecture: Defined in model.py.
3. Training Loop: Defined in the notebook with performance metrics.

Evaluation

The trained model is evaluated on a validation set, and the best-performing model is saved. The evaluation metrics include accuracy and loss.

Streamlit App

A Streamlit application has been developed to facilitate the deployment of the model and enable predictions on new MRI images. The app can be accessed here.

Functionality:

1. Model Loading: The pre-trained model is loaded automatically upon accessing the app.
2. Image Upload: Users can upload MRI images directly to the app interface.
3. Prediction Display: Once an image is uploaded, the app displays the predicted tumor type based on the model's classification.

The Streamlit app provides a user-friendly interface for interacting with the model and obtaining predictions effortlessly.

Run the Streamlit app:

streamlit run app.py

Project Demo

Project.Demo.mp4

Usage

1. Clone the repository:

git clone https://github.com/HalemoGPA/BrainMRI-Tumor-Classifier-Pytorch.git
cd BrainMRI-Tumor-Classifier-Pytorch

2. Install dependencies:

pip install -r requirements.txt

3. Run the Streamlit app:

streamlit run app.py

Results

The model achieves an accuracy of 99.3% on the test set. Training and validation loss and accuracy plots are provided to visualize the model's performance. Confusion matrices illustrate the classification performance on the test set.

Acknowledgments

• The dataset is provided by MASOUD NICKPARVAR on Kaggle.
• This project uses PyTorch for building and training the model.

Kaggle Notebook

The model training and evaluation process is detailed in this Kaggle Notebook.