Error while processing tag 00291003

[Girosella]

I just encountered a runtime error: Error while processing tag 00291003
judging from the error context, it seems some of my data got the wrong meta values, tag 00291003, I downloaded one good and bad case separately, while in Horos, no tag named 00291003 was found.
is there anyone can help?

config:
MONAI version: 1.0.1

here comes the complete error info:

UserWarning: Invalid value for VR IS: 'F'. Please see <https://dicom.nema.org/medical/dicom/current/output/html/part05.html#table_6.2-1> for allowed values for each VR.
  warnings.warn(msg)
---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:91, in apply_transform(transform, data, map_items, unpack_items, log_stats)
     90         return [_apply_transform(transform, item, unpack_items) for item in data]
---> 91     return _apply_transform(transform, data, unpack_items)
     92 except Exception as e:
     93     # if in debug mode, don't swallow exception so that the breakpoint
     94     # appears where the exception was raised.

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:55, in _apply_transform(transform, parameters, unpack_parameters)
     53     return transform(*parameters)
---> 55 return transform(parameters)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/io/dictionary.py:154, in LoadImaged.__call__(self, data, reader)
    153 for key, meta_key, meta_key_postfix in self.key_iterator(d, self.meta_keys, self.meta_key_postfix):
--> 154     data = self._loader(d[key], reader)
    155     if self._loader.image_only:

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/io/array.py:266, in LoadImage.__call__(self, filename, reader)
    265 img_array: NdarrayOrTensor
--> 266 img_array, meta_data = reader.get_data(img)
    267 img_array = convert_to_dst_type(img_array, dst=img_array, dtype=self.dtype)[0]

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/data/image_reader.py:589, in PydicomReader.get_data(self, data)
    588 if not isinstance(data[0], List):
--> 589     dicom_data.append(self._combine_dicom_series(data))
    590 # a list of list, each inner list represents a dicom series
    591 else:

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/data/image_reader.py:548, in PydicomReader._combine_dicom_series(self, data)
    547 stack_array = np.stack(stack_array, axis=-1)
--> 548 stack_metadata = self._get_meta_dict(first_slice)
    549 stack_metadata["spacing"] = np.asarray(spacing)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/data/image_reader.py:644, in PydicomReader._get_meta_dict(self, img)
    636 """
    637 Get all the metadata of the image and convert to dict type.
    638 
   (...)
    641 
    642 """
--> 644 metadata = img.to_json_dict()
    646 if self.prune_metadata:

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/pydicom/dataset.py:2504, in Dataset.to_json_dict(self, bulk_data_threshold, bulk_data_element_handler, suppress_invalid_tags)
   2503         if not suppress_invalid_tags:
-> 2504             raise exc
   2506 return json_dataset

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/pydicom/dataset.py:2497, in Dataset.to_json_dict(self, bulk_data_threshold, bulk_data_element_handler, suppress_invalid_tags)
   2496 try:
-> 2497     json_dataset[json_key] = data_element.to_json_dict(
   2498         bulk_data_element_handler=bulk_data_element_handler,
   2499         bulk_data_threshold=bulk_data_threshold
   2500     )
   2501 except Exception as exc:

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/pydicom/dataelem.py:377, in DataElement.to_json_dict(self, bulk_data_element_handler, bulk_data_threshold)
    376 if 'Value' in json_element:
--> 377     json_element['Value'] = jsonrep.convert_to_python_number(
    378         json_element['Value'], self.VR
    379     )
    380 return json_element

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/pydicom/jsonrep.py:61, in convert_to_python_number(value, vr)
     60 if isinstance(value, (list, tuple)):
---> 61     return [
     62         number_type(v) if v is not None
     63         else empty_value_for_VR(vr)
     64         for v in value
     65     ]
     67 return number_type(value)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/pydicom/jsonrep.py:62, in <listcomp>(.0)
     60 if isinstance(value, (list, tuple)):
     61     return [
---> 62         number_type(v) if v is not None
     63         else empty_value_for_VR(vr)
     64         for v in value
     65     ]
     67 return number_type(value)

ValueError: invalid literal for int() with base 10: 'F'

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

RuntimeError                              Traceback (most recent call last)
File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:91, in apply_transform(transform, data, map_items, unpack_items, log_stats)
     90         return [_apply_transform(transform, item, unpack_items) for item in data]
---> 91     return _apply_transform(transform, data, unpack_items)
     92 except Exception as e:
     93     # if in debug mode, don't swallow exception so that the breakpoint
     94     # appears where the exception was raised.

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:55, in _apply_transform(transform, parameters, unpack_parameters)
     53     return transform(*parameters)
---> 55 return transform(parameters)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/compose.py:174, in Compose.__call__(self, input_)
    173 for _transform in self.transforms:
--> 174     input_ = apply_transform(_transform, input_, self.map_items, self.unpack_items, self.log_stats)
    175 return input_

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:118, in apply_transform(transform, data, map_items, unpack_items, log_stats)
    117         _log_stats(data=data)
--> 118 raise RuntimeError(f"applying transform {transform}") from e

RuntimeError: applying transform <monai.transforms.io.dictionary.LoadImaged object at 0x7f582d8600d0>

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

RuntimeError                              Traceback (most recent call last)
Input In [38], in <cell line: 12>()
     16 epoch_loss = 0
     17 step = 0
---> 19 for batch_data in train_loader:
     20     try:
     21         inputs, msks, labels = batch_data['img'].to(device),batch_data['msk'].to(device),batch_data['label'].to(device)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/torch/utils/data/dataloader.py:530, in _BaseDataLoaderIter.__next__(self)
    528 if self._sampler_iter is None:
    529     self._reset()
--> 530 data = self._next_data()
    531 self._num_yielded += 1
    532 if self._dataset_kind == _DatasetKind.Iterable and \
    533         self._IterableDataset_len_called is not None and \
    534         self._num_yielded > self._IterableDataset_len_called:

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/torch/utils/data/dataloader.py:570, in _SingleProcessDataLoaderIter._next_data(self)
    568 def _next_data(self):
    569     index = self._next_index()  # may raise StopIteration
--> 570     data = self._dataset_fetcher.fetch(index)  # may raise StopIteration
    571     if self._pin_memory:
    572         data = _utils.pin_memory.pin_memory(data)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py:49, in _MapDatasetFetcher.fetch(self, possibly_batched_index)
     47 def fetch(self, possibly_batched_index):
     48     if self.auto_collation:
---> 49         data = [self.dataset[idx] for idx in possibly_batched_index]
     50     else:
     51         data = self.dataset[possibly_batched_index]

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/torch/utils/data/_utils/fetch.py:49, in <listcomp>(.0)
     47 def fetch(self, possibly_batched_index):
     48     if self.auto_collation:
---> 49         data = [self.dataset[idx] for idx in possibly_batched_index]
     50     else:
     51         data = self.dataset[possibly_batched_index]

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/data/dataset.py:104, in Dataset.__getitem__(self, index)
    101 if isinstance(index, collections.abc.Sequence):
    102     # dataset[[1, 3, 4]]
    103     return Subset(dataset=self, indices=index)
--> 104 return self._transform(index)

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/data/dataset.py:90, in Dataset._transform(self, index)
     86 """
     87 Fetch single data item from `self.data`.
     88 """
     89 data_i = self.data[index]
---> 90 return apply_transform(self.transform, data_i) if self.transform is not None else data_i

File ~/anaconda3/envs/radiomics_gyy/lib/python3.10/site-packages/monai/transforms/transform.py:118, in apply_transform(transform, data, map_items, unpack_items, log_stats)
    116     else:
    117         _log_stats(data=data)
--> 118 raise RuntimeError(f"applying transform {transform}") from e

RuntimeError: applying transform <monai.transforms.compose.Compose object at 0x7f582d863850>

[wyli]

MONAI/monai/data/image_reader.py

Line 650 in 78d4f42

metadata = img.to_json_dict()

this function call may take an additional arg:

metadata = img.to_json_dict(suppress_invalid_tags=True)

could you please help verify this? or send us an example image for debugging. cc @yiheng-wang-nv

[Girosella]

OK，I would try that and add the result in next comment.
I checked the problem image, it's all right to load and read using first(dataloader) -like codes but not ok when training.

[Girosella]

MONAI/monai/data/image_reader.py

Line 650 in 78d4f42

metadata = img.to_json_dict()

this function call may take an additional arg:

metadata = img.to_json_dict(suppress_invalid_tags=True)

could you please help verify this? or send us an example image for debugging. cc @yiheng-wang-nv

sorry for the delay, I changed the dcm to nii and trained my model. I just finished the training and tried your suggestion.
changed the image_reader.py, relaunch the kernel, and prepare the data and so on.
Here is what I found.

I suppose first time it got problem in cachedatset when dealing with dcm files but second time it's all right.
the training is successful and done.
I also tried with dataset, the training is also ok. but dataset trial is after cachedataset, so I am not sure this is not effective since I've already loaded the data in cache.
But anyway, your suggestion seems working.
Thanks.