We generated those 2 files ourselves from BRATS 2013 dataset. We use synthetic flair images particularly and saved them with there ground truth masks in numpy files. “BRATS2013_Syn_Flair_Train_X.npy” contains input images and “BRATS2013_Syn_Flair_Train_S.npy” contains segmentation masks only.
We generated those 2 files ourselves from BRATS 2013 dataset. We use synthetic flair images particularly and saved them with there ground truth masks in numpy files. “BRATS2013_Syn_Flair_Train_X.npy” contains input images and “BRATS2013_Syn_Flair_Train_S.npy” contains segmentation masks only.
Originally posted by @mahfuzmohammad in https://github.com/MrGiovanni/UNetPlusPlus/issues/12#issuecomment-485226086
So could you provide the preprocess code? from BRATS 2013 dataset to the images and labels?