Save 3D segmentation results of each channel directly

[YingdiZhang]

Thanks for the great job. I am using monai in a 3D segmentation task with output channel 6. I want to save the output of each channel directly. How should I write the postprocess and save image transform?

[KumoLiu]

Hi @YingdiZhang, you could try to use SplitDimd and SaveImaged. Here is a simple example.

trans = Compose(
    [
        LoadImaged(keys='img', image_only=True, ensure_channel_first=True),
        AsDiscreted(keys='img', to_onehot=6),
        FromMetaTensord(keys='img'),
        SplitDimd(keys='img'),
        SaveImaged(keys=['img_0', 'img_1'], output_dir='.', )
])

Hope it can help you, thanks!

[YingdiZhang]

Thanks for your reply. I use monai 0.9.0 and write a save transform myself, but it occurs an error, and I dont know why. My code is:

self.post_trans_save = Compose(
[
EnsureTyped(keys="heatmap_pred"), #
Invertd(
keys="heatmap_pred", # invert the heatmap_pred data field, also support multiple fields
transform=self.transform,
orig_keys="image",
# get the previously applied self.transforms information on the image data field,
# then invert pred based on this information. we can use same info
# for multiple fields, also support different orig_keys for different fields
meta_keys="heatmap_pred_meta_dict", #"heatmap_pred_meta_dict", # # "heatmap_pred_L1_meta_dict",
# key field to save inverted meta data, every item maps to keys
orig_meta_keys="image_meta_dict", # get the meta data from image_meta_dict field when inverting,
# for example, may need the affine to invert Spacingd transform,
# multiple fields can use the same meta data to invert
meta_key_postfix="meta_dict",
# if meta_keys=None, use "{keys}{meta_key_postfix}" as the meta key,
# if orig_meta_keys=None, use "{orig_keys}{meta_key_postfix}",
# otherwise, no need this arg during inverting
nearest_interp=False, # don't change the interpolation mode to "nearest" when inverting transforms
# to ensure a smooth output, then execute AsDiscreted transform
to_tensor=True # convert to PyTorch Tensor after inverting
),
# AsDiscreted(keys="pred", argmax=True),
SplitChanneld(keys="heatmap_pred"),
SaveImaged(keys="heatmap_pred_0", meta_keys="heatmap_pred_meta_dict", output_dir=self.pred_output_folder,
output_postfix="heatmap_L1", resample=False, output_ext='.nii.gz',
separate_folder=False)
]
)

The error is:
transform info of image is not available or no InvertibleTransform applied.
0%| | 0/18 [00:05<?, ?it/s]
Traceback (most recent call last):
File "/home/zyd/anaconda3/envs/med/lib/python3.9/site-packages/monai/transforms/transform.py", line 89, in apply_transform
return _apply_transform(transform, data, unpack_items)
File "/home/zyd/anaconda3/envs/med/lib/python3.9/site-packages/monai/transforms/transform.py", line 53, in _apply_transform
return transform(parameters)
File "/home/zyd/anaconda3/envs/med/lib/python3.9/site-packages/monai/transforms/io/dictionary.py", line 271, in call
meta_data = d[meta_key] if meta_key is not None else None
KeyError: 'heatmap_pred_meta_dict'

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

Traceback (most recent call last):
File "/data/zyd/beifen_xinsuoqu_/MedSeg/MedSeg/test/Task960Tester.py", line 663, in
t960_tester.test(
File "/data/zyd/beifen_xinsuoqu_/MedSeg/MedSeg/test/Task960Tester.py", line 401, in test
_ = [self.post_trans_save_heatmap_L1(i) for i in decollate_batch(val_data)]
File "/data/zyd/beifen_xinsuoqu_/MedSeg/MedSeg/test/Task960Tester.py", line 401, in
_ = [self.post_trans_save_heatmap_L1(i) for i in decollate_batch(val_data)]
File "/home/zyd/anaconda3/envs/med/lib/python3.9/site-packages/monai/transforms/compose.py", line 173, in call
input_ = apply_transform(transform, input, self.map_items, self.unpack_items, self.log_stats)
File "/home/zyd/anaconda3/envs/med/lib/python3.9/site-packages/monai/transforms/transform.py", line 113, in apply_transform
raise RuntimeError(f"applying transform {transform}") from e
RuntimeError: applying transform <monai.transforms.io.dictionary.SaveImaged object at 0x7f6b8c1f4b50>

[KumoLiu]

From the error message, seems you don't have heatmap_pred_meta_dict, maybe you could comment the SaveImaged to check what meta_dict you have in your output. Thanks!