Brain Tumor Detection with MRI Images using PyTorch CNN

This project aims to detect brain tumors in MRI images using a Convolutional Neural Network (CNN) implemented in PyTorch.

Overview

The CNN model is trained on a dataset of MRI images to classify images as either containing a tumor or not. The model achieves an accuracy of over 90%.

Features

• Implementation of a CNN model in PyTorch for brain tumor detection
• Training and evaluation of the model using MRI images
• Visualization of feature maps of convolutional filters

Installation

1. Clone the repository: git clone https://github.com/Firojpaudel/Brain_Tumor_Detection.git
2. Also you'll need to install pytorch and would suggest to create an environment
3. And Install the required packages: pip install -r requirements.txt inside that environment

   I used conda

Usage

1. Run jupyter notebook in the project directory
2. Open and run the Brain_tumor_detection.ipynb notebook
3. Follow the instructions in the notebook to train and test the CNN model

Next Steps

• Address overfitting concerns due to the complexity of the model and limited dataset size
• Explore techniques such as data augmentation, regularization, or using a simpler model architecture