Negative strides issue in monai.transforms.LoadImage causes error during collate

[KumoLiu]

Describe the bug When using monai.transforms.LoadImage with the nrrdreader to load an NRRD file, the resulting image metadata contains negative strides. This negative strides issue leads to an error when the tensor is collated.

To Reproduce

import monai

# Load an NRRD file
img = monai.transforms.LoadImage(reader="nrrdreader", index_order="C")("wrong_something.nrrd")

print(img.meta['spatial_shape'].strides)  # Output: (-8,)

Error log:

torch/utils/data/_utils/collate.py", line 222, in <listcomp>
    return collate([torch.as_tensor(b) for b in batch], collate_fn_map=collate_fn_map)
ValueError: At least one stride in the given numpy array is negative, and tensors with negative strides are not currently supported. (You can probably work around this by making a copy of your array  with array.copy().)

Test data: https://drive.google.com/file/d/1Et-khdsou7E2T_aoX0zLvnld49Xh8YDn/view?usp=drive_link