Error when using ResEncUNet with nnUNetTrainer_DASegOrd0_NoMirroring

[wasserth]

When trying to train a model with plan nnUNetResEncUNetMPlans and trainer nnUNetTrainer_DASegOrd0_NoMirroring I get the following error:

Traceback (most recent call last):
  File "/mnt/nor/miniconda3/envs/nnunet_latest/lib/python3.10/site-packages/batchgenerators/dataloading/nondet_multi_threaded_augmenter.py", line 55, in producer
    item = transform(**item)
  File "/mnt/nor/miniconda3/envs/nnunet_latest/lib/python3.10/site-packages/batchgeneratorsv2/transforms/base/basic_transform.py", line 18, in __call__
    return self.apply(data_dict, **params)
  File "/mnt/nor/miniconda3/envs/nnunet_latest/lib/python3.10/site-packages/batchgeneratorsv2/transforms/utils/compose.py", line 13, in apply
    data_dict = t(**data_dict)
  File "/mnt/nor/miniconda3/envs/nnunet_latest/lib/python3.10/site-packages/batchgeneratorsv2/transforms/base/basic_transform.py", line 17, in __call__
    params = self.get_parameters(**data_dict)
  File "/mnt/nor/miniconda3/envs/nnunet_latest/lib/python3.10/site-packages/batchgeneratorsv2/transforms/spatial/spatial.py", line 58, in get_parameters
    dim = data_dict['image'].ndim - 1
KeyError: 'image'

It does not happend with trainer nnUNetTrainerNoMirroring or nnUNetTrainer_DASegOrd0.

[fitzjalen]

#2480
Same issue