Medical Zoo Pytorch

Our goal is to implementent an open-source medical image segmentation library of state of the art 2D/3D deep neural networks in PyTorch along with data loaders of the most common medical MRI datasets. The first stable release of our repository is expected to be publised in the end of September. We strongly believe in open and reproducible deep learning research. In order to reproduce our results, the code (alpha release) and materials of this work are available in this repository. This project started as an MSc Thesis and is currently under further development.

Alpha release - work in progress

For 3D multi-modal brain MRI segmentation check the thesis branch of this repository. Although this work was initially focused on 3D multi-modal brain MRI segmentation we will slowly add more architectures and dataloaders.

Top priorities

1. batch size support!!!!
2. on the fly volume vizualization in tensorboard
3. fix Brats2018 and mrbrains dataloaders
4. add hyper densenet and unet++ 3d

Documentation

1. Docker image for installation guide
2. Examples

Implemented dataloaders

ISEG 2017

MRBRAINS2018

Implemented architectures

Densenet3D (1-stream, 2-stream and 3-stream)

Unet3D

Vnet + lighter version

Support

If you like this repo and find it useful, please consider (★) starring it, so that it can reach a broader audience.