https://github.com/aadityaza/3d-uet-for-brain-mri-segmentation

https://github.com/aadityaza/3d-uet-for-brain-mri-segmentation

3d 3d-segmentation brain brats2020 dice-loss focal-loss iou pytorch segmentation tumor-segmentation unet

Last synced: 12 days ago
JSON representation

• Host: GitHub
• URL: https://github.com/aadityaza/3d-uet-for-brain-mri-segmentation
• Owner: Aadityaza
• Created: 2024-08-20T06:27:39.000Z (3 months ago)
• Default Branch: main
• Last Pushed: 2024-08-26T19:25:39.000Z (3 months ago)
• Last Synced: 2024-10-17T12:37:41.730Z (29 days ago)
• Topics: 3d, 3d-segmentation, brain, brats2020, dice-loss, focal-loss, iou, pytorch, segmentation, tumor-segmentation, unet
• Language: Jupyter Notebook
• Homepage:
• Size: 40.4 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Brain Tumor Segmentation Using 3D-Unet 

![image](https://github.com/user-attachments/assets/fe1509e6-b736-40da-9e91-fd13746a6d55)

## Overview

This project implements a deep learning model for brain tumor segmentation using the BraTS2020 dataset. The model processes multi-modal MRI scans to identify and segment different types of brain tumors.

## Data

The project uses the BraTS2020_TrainingData dataset, which includes the following MRI modalities:

- T1-weighted (T1)

- T2-weighted (T2)

- T1 with contrast enhancement (T1ce)

- Fluid Attenuated Inversion Recovery (FLAIR)

## Requirements

All required dependencies are listed in the `requirements.txt` file. To install them, run:

```

pip install -r requirements.txt

```

## Data Preprocessing

The preprocessing pipeline includes the following steps:

1. Loading and scaling of multi-modal MRI images (T1, T2, T1ce, FLAIR)

2. Loading and processing of segmentation masks

3. Combining multiple modalities into a single 4-channel image

4. Cropping images and masks to a specified region of interest

5. One-hot encoding of segmentation masks

6. Conversion to PyTorch tensors

7. Saving processed data as PyTorch tensor files

## Dataset Class

The `BrainTumorDataset` class is implemented to facilitate data loading during training. It inherits from PyTorch's `Dataset` class and provides methods for accessing the preprocessed image and mask pairs.

## Model Training

The training process includes:

1. Data loading using `DataLoader`

2. Model optimization using a combination of Dice loss and Focal loss

3. Training and validation loops

4. Logging of training and validation losses

## Usage

1. Preprocess the BraTS2020 data using the provided preprocessing script.

2. Create instances of the `BrainTumorDataset` class for training and validation sets.

3. Initialize the model, optimizer, and loss functions.

4. Run the training loop, calling `train_epoch` and `validate` functions for each epoch.

## Note

This project is based on BraTS2020 dataset. Ensure you have the correct data structure and file paths before running the preprocessing and training scripts.