file name extraction

[sepidhk]

hi all,
in the previous version of MONAI, I use the
fname= os.path.basename(test_data["img_meta_dict"]["filename_or_obj"][0])
to extract the file name from data dictionary and it seems like this key is not available now. I would appreciate it if you could advice how to get the file name.

[KumoLiu]

Hi @sepidhk, use test_data.meta["filename_or_obj"] instead.

We introduced MetaTensor since 0.9.0: https://github.com/Project-MONAI/MONAI/wiki/MetaTensor-guide

Thanks.

[sepidhk]

get the following error

`---------------------------------------------------------------------------
AttributeError Traceback (most recent call last)
Cell In[71], line 23
20 predicted = torch.argmax(test_data["pred"], dim=1, keepdim=True).detach().cpu()
21 predicted_mask= np.moveaxis(np.squeeze(predicted, 0), 0, -1)
---> 23 filename = test_data.meta["filename_or_obj"]
25 predicted_mask[predicted_mask == 0] = 0
26 predicted_mask[predicted_mask == 1] = 53

AttributeError: 'dict' object has no attribute 'meta'`

[sepidhk]

here is my data transform, reading the images using PILReader:
test_transforms_encoded = Compose([
LoadImaged(keys=["img", "label"], reader=PILReader),
EnsureChannelFirstd(keys=["img", "label"]),
ScaleIntensityd(keys=["img"], allow_missing_keys=True),
Flipd(keys=["img", "label"], spatial_axis=0),
Rotate90d(keys=["img", "label"], k=3),
EnsureTyped(keys=["img", "label"]),
AsDiscreted(keys=["label"], to_onehot=num_classes),
])

test_ds_encoded = monai.data.Dataset(data=val_files, transform=test_transforms_encoded)
test_loader_encoded = DataLoader(test_ds_encoded, batch_size=1, shuffle=False, num_workers=1)

[sepidhk]

Solved by:
filename = test_inputs.meta["filename_or_obj"]