:param normalization_scheme_per_modality: dict {0:'nonCT'} :param use_nonzero_mask: {0:False} :param intensityproperties:

This one finds a mask of nonzero elements (must be nonzero in all modalities) and crops the image to that mask. In the case of BRaTS

This is the basic data loader for 2D networks. It uses preprocessed data as produced by my (Fabian) preprocessing. You can load the d

A generic class that can train almost any neural network (RNNs excluded). It provides basic functionality such as the training loop,

:param deterministic: :param fold: can be either [0 ... 5) for cross-validation, 'all' to train on all available training data or

:param dont_do_if_covers_more_than_X_percent: dont_do_if_covers_more_than_X_percent=0.25 is 25\%! :param channel_idx: can be list or

2019_11_22: I have no idea what the purpose of this was... the same as above but here we should use only expanding operations. Expan

data[mask < 0] = 0 Sets everything outside the mask to 0. CAREFUL! outside is defined as < 0, not =0 (in the Mask)!!! :param

regions are tuple of tuples where each inner tuple holds the class indices that are merged into one region, example: regions= ((1, 2)

based on matthews correlation coefficient https://en.wikipedia.org/wiki/Matthews_correlation_coefficient Does not work. Real

squares the terms in the denominator as proposed by Milletari et al.

CAREFUL. Weights for CE and Dice do not need to sum to one. You can set whatever you want. :param soft_dice_kwargs: :param ce

DO NOT APPLY NONLINEARITY IN YOUR NETWORK! THIS LOSS IS INTENDED TO BE USED FOR BRATS REGIONS ONLY :param soft_dice_kwargs:

use this if you have several outputs and ground truth (both list of same len) and the loss should be computed between them (x[0] and

Following UNet building blocks can be added by utilizing the properties this class exposes (TODO) this one includes the bottleneck l

Following UNet building blocks can be added by utilizing the properties this class exposes (TODO) this one includes the bottleneck l