EnhGAN

Enhancement GAN

We present a novel architecture based on conditional generative adversarial networks (cGANs) to improve the lesion contrast for segmentation.

You can find detailed results (Team name: Hamghalam) on BraTS 2019 dataset on:

- Validation Phase Leaderboard 2019 -

https://www.cbica.upenn.edu/BraTS19/lboardValidation.html

- Training Phase Leaderboard 2019 -

https://www.cbica.upenn.edu/BraTS19/lboardTraining.html

High Tissue Contrast MRI Synthesis

Prerequisites

A CUDA compatable GPU with memory not less than 12GB is recommended for training. For testing only, a smaller GPU should be suitable.

Linux or OSX

NVIDIA GPU + CUDA CuDNN

Keras

SimpleITK

Pretrained model

Download pretrained model (trained on BraTS dataset.) on this address:

https://drive.google.com/open?id=1Gc-gbrq-KoI67tgn-nFiCSdY5jd0dO4y

Prepare dataset

1- Put your dataset (here BraTS) on the root address:

2- Create "data_adr.txt" file and determine requirement as bellow:

#############################################

[data]

data_root = /home/mohammad/input/MICCAI_BraTS17_Data_Training/

data_names = config/train_name_all.txt

modality_postfix = [flair]

file_postfix = nii.gz

#############################################

Put name of each Subject ID on "train_name_all.txt"

To compute high contrast images based on EnhGAN model, run this command on Linux terminal:

python Enhancement_GAN.py config/data_adr.txt

How to download data

BraTS 2019 dataset. Data can be downloaded from http://braintumorsegmentation.org/