Version corresponding to the article published on arXiv in April 2024. Attached are the model weights as PyTorch state_dicts. The filename specifies the domain used for training and whether the weights performed the best on the respective domain or whether it performed close to the average performance on the respective domain. The *.npy files contain the normalization parameters (mean and standard deviation) used for input channel normalization (cf. function compute_normalized_patch(...) in ./datasetsSegmentationDatasetPatchProviders.py lines 83 ff.).