Image and label orientation using Orientationd transform

[Suchi97]

Hi Team,

We used a sample image and the corresponding label (BTCV dataset : img0001.nii.gz and label0001.nii.gz) to test the orientationd transform.

Below transforms are used for the above purpose.

transforms = Compose([LoadImaged(keys=["image", "label"]), EnsureChannelFirstd(keys=["image","label"]),
                           ScaleIntensityRanged(keys=["image"], a_min=-175, a_max=250, b_min=0, b_max=1.0, clip=True),
                           Orientationd(keys=["image","label"], axcodes='RAS'),    # axcodes = RAS, RPI, LPI, LAS
                           EnsureTyped(keys=["image","label"])
                          ])

When we used the orientationd, we changed the axcodes values and for each axcodes, we obtained the below results.

RAS

RPI

LPI

LAS

From the above images, we can see that the images and labels are not aligned properly.

Can you please help me to solve this issue.
Thanks in advance

[rijobro]

Can you:

1. Show what the images look like with just LoadImage(["image", "label"])?
2. Print image_meta_dict and label_meta_dict (of just the difference between them)?
3. Show what code you're using for the matplotlib?