Post Transform Invertd error

[jonathanquantumalpha]

Hi I've created a simple app which segment tumors for colon. I want to apply post_transforms for the original image (instead of the transformed) as part of the final output shown, however my invertd came to the following error:

RuntimeError: applying transform <monai.transforms.post.dictionary.Invertd object at 0x2cefdb520>
Traceback:
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 542, in _run_script
exec(code, module.dict)
File "/Users/Python_Projects/post.py", line 157, in
test_outputs = post_transforms({"image": test_images, "pred": test_outputs})
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/compose.py", line 322, in call
result = execute_compose(
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/compose.py", line 111, in execute_compose
data = apply_transform(
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 171, in apply_transform
raise RuntimeError(f"applying transform {transform}") from e

Here is the code snippet:

                    # Assuming test_data contains the original images and test_outputs contains the predictions
                    with torch.no_grad():
                        for test_batch in test_org_loader:
                            test_images = test_batch["image"].to(device)
                            roi_size = (128, 128, 64)
                            sw_batch_size = 4
                            test_outputs = sliding_window_inference(test_images, roi_size, sw_batch_size, model)
                            test_outputs = post_transforms({"image": test_images, "pred": test_outputs})

                            # uncomment the following lines to visualize the predicted results
                            test_output = from_engine(["pred"])(test_data)


                            original_image = loader(test_output[0].meta["filename_or_obj"])

                            # Load the NIfTI file to determine the number of slices
                            loader = LoadImage(image_only=True)
                            nifti_data = loader(file_path)
                            num_slices = nifti_data.shape[-1]  # Assuming the slices are along the last dimension

                            for j in range(num_slices): 
                                if j < test_data["image"].shape[4]:
                                    fig = plt.figure("check", (12, 4))
                                    plt.subplot(1, 2, 1)
                                    plt.title(f"image {j}")
                                    plt.imshow(original_image["image"][0, 0, :, :, j], cmap="gray")
                                    plt.subplot(1, 2, 2)
                                    plt.title(f"output {j}")
                                    plt.imshow(torch.argmax(test_outputs, dim=1).detach().cpu()[0, :, :, j] == 1)
                                    st.pyplot(fig)

            st.success('Nifti files processed successfuly!')


else:
    st.warning('Please upload Nifti files or a zipped folder.')

[KumoLiu]

Hi @jonathanquantumalpha, could you please paste the whole error message?
BTW, you could refer to this tutorial.
https://github.com/Project-MONAI/tutorials/blob/main/modules/inverse_transforms_and_test_time_augmentations.ipynb

Hope it helps, thanks.

[jonathanquantumalpha]

Here it is:

Traceback (most recent call last):
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 141, in apply_transform
return _apply_transform(transform, data, unpack_items, lazy, overrides, log_stats)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 98, in _apply_transform
return transform(data, lazy=lazy) if isinstance(transform, LazyTrait) else transform(data)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/spatial/dictionary.py", line 857, in inverse
d[key] = self.resizer.inverse(d[key])
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/spatial/array.py", line 851, in inverse
transform = self.pop_transform(data)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/inverse.py", line 328, in pop_transform
return self.get_most_recent_transform(data, key, check, pop=True)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/inverse.py", line 310, in get_most_recent_transform
self.check_transforms_match(all_transforms[-1])
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/inverse.py", line 264, in check_transforms_match
xform_id = transform.get(TraceKeys.ID, "")
AttributeError: 'list' object has no attribute 'get'

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

Traceback (most recent call last):
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 141, in apply_transform
return _apply_transform(transform, data, unpack_items, lazy, overrides, log_stats)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 98, in _apply_transform
return transform(data, lazy=lazy) if isinstance(transform, LazyTrait) else transform(data)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/post/dictionary.py", line 702, in call
inverted = self.transform.inverse(input_dict)
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/compose.py", line 351, in inverse
data = apply_transform(
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 171, in apply_transform
raise RuntimeError(f"applying transform {transform}") from e
RuntimeError: applying transform <bound method Resized.inverse of <monai.transforms.spatial.dictionary.Resized object at 0x16df58580>>

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

Traceback (most recent call last):
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/streamlit/runtime/scriptrunner/script_runner.py", line 542, in _run_script
exec(code, module.dict)
File "/Users/Python_Projects/EdenSehat/deploy/post.py", line 157, in
test_outputs = post_transforms({"image": test_images, "pred": test_outputs})
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/compose.py", line 322, in call
result = execute_compose(
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/compose.py", line 111, in execute_compose
data = apply_transform(
File "/Users/miniconda3/envs/5organs/lib/python3.10/site-packages/monai/transforms/transform.py", line 171, in apply_transform
raise RuntimeError(f"applying transform {transform}") from e
RuntimeError: applying transform <monai.transforms.post.dictionary.Invertd object at 0x16df5b850>

[KumoLiu]

From the code snippet you've provided, it seems you're applying a post-transform to a batch of data. Could you please try integrating the decollate_batch function? Here's an example:

for val_data in val_org_loader:
    val_inputs = val_data["image"].to(device)
    val_data["pred"] = inference(val_inputs)
    val_data = [post_transforms(i) for i in decollate_batch(val_data)]

I hope this proves helpful. Thank you!

[jonathanquantumalpha]

[KumoLiu]

Could you please provide a small piece of code that can reproduce the issue? Without it, pinpointing the problem may prove challenging. However, I would recommend starting with the notebook first. After successfully running it, you can make minor changes to accommodate your own data.
Here's the link to the notebook: https://github.com/Project-MONAI/tutorials/blob/main/3d_segmentation/brats_segmentation_3d.ipynb.