Brain Tumor Segmentation using U-Net

A deep learning project for brain tumor segmentation from MRI images using the U-Net architecture. This repository includes data preprocessing, model training, evaluation, and inference scripts.

📌 Features

• U-Net implementation in PyTorch
• MRI image preprocessing
• Automatic dataset splitting
• Model training and validation
• Dice Score and IoU evaluation
• Model checkpoint saving
• Inference on new MRI images

🛠️ Tech Stack

• Python 3.11
• PyTorch
• NumPy
• OpenCV
• Albumentations
• Matplotlib
• scikit-learn

📊 Dataset

This project is designed for brain MRI segmentation datasets such as:

• BraTS 2021
• BraTS 2023

Download the dataset from the official BraTS website.

🚀 Installation

Clone the repository

git clone https://github.com/yourusername/your-repository.git
cd your-repository

Install dependencies

pip install -r requirements.txt

⚙️ Data Preprocessing

python preprocess.py

🏋️ Train the Model

python train.py

📈 Evaluate

python evaluate.py

🔍 Inference

python inference.py

📐 Evaluation Metrics

• Dice Score
• Intersection over Union (IoU)
• Loss Curve

🖼️ Sample Results

MRI Image	Ground Truth	Prediction
Coming Soon	Coming Soon	Coming Soon

📦 Requirements

• Python 3.11+
• PyTorch
• CUDA (Optional)

Install:

pip install -r requirements.txt

🤝 Contributing

Contributions are welcome! Feel free to open an issue or submit a pull request.

📄 License

This project is licensed under the MIT License.

👨‍💻 Author

Md. Arafat Hossain

GitHub: https://github.com/Arafat-shuvo

---

If you find this project useful, please ⭐ the repository.