This is the code for my graduation thesis and the solution for my BRATS 2020 Challenge.

Here are the four modes of the image and its mask

First,the image is cropped into patches.Runsrc/partition.py,change the storage address of training set, verification set and test set successively.

train_brats_path = "your-Path_to/MICCAI_BraTS_2020_Data_Training"
output_trainImage = "your-Path_to/trainImage"
output_trainMask = "your-Path_to/trainMask"

And runsrc/train.py，change data_path.Began to run

data_path = 'your-Path_to'

Runsrc/inference.py，to restore the picture.

Runsrc/test.py.Test it.

训练集对比 训练集对比 验证集对比

训练集评价指标

验证集评价指标