A RuntimeError raised by torchIO

[lyangfan]

Hi:
I got a RuntimeError raised by torchIO:

RuntimeError                              Traceback (most recent call last)
/usr/local/lib/python3.7/dist-packages/torchio/data/subject.py in check_consistent_space(self)
    308             self.check_consistent_attribute('origin')
--> 309             self.check_consistent_spatial_shape()
    310         except RuntimeError as e:

14 frames
/usr/local/lib/python3.7/dist-packages/torchio/data/subject.py in check_consistent_spatial_shape(self)
    295     def check_consistent_spatial_shape(self) -> None:
--> 296         self.check_consistent_attribute('spatial_shape')
    297 

/usr/local/lib/python3.7/dist-packages/torchio/data/subject.py in check_consistent_attribute(self, attribute, relative_tolerance, absolute_tolerance, message)
    283                     )
--> 284                     raise RuntimeError(message)
    285         except TypeError:

RuntimeError: More than one value for "spatial_shape" found in subject images:
{'label': (400, 400, 217), 'source': (400, 400, 218)}

The above exception was the direct cause of the following exception:

RuntimeError                              Traceback (most recent call last)
<ipython-input-20-4dc2ba0c028a> in <module>()
----> 1 train_model()

<ipython-input-19-451b42dea421> in train_model()
     45 
     46     trainLosses, valLosses = train_valid.train(numEpoch, patchesTrainingLoader,
---> 47         patchesValLoader, model, optimizer, device)
     48 
     49     weightsPath = 'patches_state_dict.pth'

/content/train_valid.py in train(numEpochs, trainingLoader, validationLoader, model, optimizer, device)
     37     trainLosses = []
     38     valLosses = []
---> 39     valLosses.append(runEpoch(0, 'valid', validationLoader, model, optimizer, device))
     40     for epochIdx in range(1, numEpochs + 1):
     41         print('Starting epoch', epochIdx)

/content/train_valid.py in runEpoch(epochIdx, mode, loader, model, optimizer, device)
     16     model.to(device)
     17 
---> 18     for batchIdx, batch in enumerate(tqdm(loader)):
     19         inputs, targets = prepareBatch(batch, device)
     20         optimizer.zero_grad()

/usr/local/lib/python3.7/dist-packages/tqdm/std.py in __iter__(self)
   1178 
   1179         try:
-> 1180             for obj in iterable:
   1181                 yield obj
   1182                 # Update and possibly print the progressbar.

/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py in __next__(self)
    519             if self._sampler_iter is None:
    520                 self._reset()
--> 521             data = self._next_data()
    522             self._num_yielded += 1
    523             if self._dataset_kind == _DatasetKind.Iterable and \

/usr/local/lib/python3.7/dist-packages/torch/utils/data/dataloader.py in _next_data(self)
    559     def _next_data(self):
    560         index = self._next_index()  # may raise StopIteration
--> 561         data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
    562         if self._pin_memory:
    563             data = _utils.pin_memory.pin_memory(data)

/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py in fetch(self, possibly_batched_index)
     42     def fetch(self, possibly_batched_index):
     43         if self.auto_collation:
---> 44             data = [self.dataset[idx] for idx in possibly_batched_index]
     45         else:
     46             data = self.dataset[possibly_batched_index]

/usr/local/lib/python3.7/dist-packages/torch/utils/data/_utils/fetch.py in <listcomp>(.0)
     42     def fetch(self, possibly_batched_index):
     43         if self.auto_collation:
---> 44             data = [self.dataset[idx] for idx in possibly_batched_index]
     45         else:
     46             data = self.dataset[possibly_batched_index]

/usr/local/lib/python3.7/dist-packages/torchio/data/queue.py in __getitem__(self, _)
    163         if not self.patches_list:
    164             self._print('Patches list is empty.')
--> 165             self._fill()
    166         sample_patch = self.patches_list.pop()
    167         self.num_sampled_patches += 1

/usr/local/lib/python3.7/dist-packages/torchio/data/queue.py in _fill(self)
    228         for _ in iterable:
    229             subject = self._get_next_subject()
--> 230             iterable = self.sampler(subject)
    231             patches = list(islice(iterable, self.samples_per_volume))
    232             self.patches_list.extend(patches)

/usr/local/lib/python3.7/dist-packages/torchio/data/sampler/sampler.py in __call__(self, subject, num_patches)
     81             num_patches: Optional[int] = None,
     82             ) -> Generator[Subject, None, None]:
---> 83         subject.check_consistent_space()
     84         if np.any(self.patch_size > subject.spatial_shape):
     85             message = (

/usr/local/lib/python3.7/dist-packages/torchio/data/subject.py in check_consistent_space(self)
    315                 ' https://github.com/fepegar/torchio/issues/647#issuecomment-913025695'
    316             )
--> 317             raise RuntimeError(message) from e
    318 
    319     def get_images_names(self) -> List[str]:

RuntimeError: As described above, some images in the subject are not in the same space. You probably can use the transforms ToCanonical and Resample to fix this, as explained at https://github.com/fepegar/torchio/issues/647#issuecomment-913025695

But the scalarImage and the LabelMap have been transformed to the same space:

      if mode == 'train':
            transforms = tio.Compose([
                ToCanonical(),
                Resample(1),
                ZNormalization(masking_method=tio.ZNormalization.mean),
                RandomAffine(),
                RandomNoise(),
                OneHot(),
            ])
        elif mode == 'validation' or 'test':
            transforms = tio.Compose([
                ToCanonical(),
                Resample(1),
                ZNormalization(masking_method=tio.ZNormalization.mean),
                OneHot(),
            ])

[fepegar]

Hi, @lyangfan. Your images seem to have slightly different original affine matrices or shapes. Can you take a look at that? You might want to resample one of them to 1 mm isotropic and then use that as reference for a second resampling transform. Does that make sense?