SwinUNETR /BTCV: transform difference in validation and test phase

[YunchaoYang]

Hi Dear MONAI researchers,

I am reading the SwinUNETR/BTCV code, I found that there is difference between in the dataset transform in the utils/data_utils.py.

For validation:

    val_transform = transforms.Compose(
        [
            transforms.LoadImaged(keys=["image", "label"]),
            transforms.AddChanneld(keys=["image", "label"]),
            transforms.Orientationd(keys=["image", "label"], axcodes="RAS"),
            transforms.Spacingd(
                keys=["image", "label"], pixdim=(args.space_x, args.space_y, args.space_z), mode=("bilinear", "nearest")
            ),
            transforms.ScaleIntensityRanged(
                keys=["image"], a_min=args.a_min, a_max=args.a_max, b_min=args.b_min, b_max=args.b_max, clip=True
            ),
            transforms.CropForegroundd(keys=["image", "label"], source_key="image"),
            transforms.ToTensord(keys=["image", "label"]),
        ]
    )

For testing:

  test_transform = transforms.Compose(
       [
           transforms.LoadImaged(keys=["image", "label"]),
           transforms.AddChanneld(keys=["image", "label"]),
           # transforms.Orientationd(keys=["image"], axcodes="RAS"),
           transforms.Spacingd(keys="image", pixdim=(args.space_x, args.space_y, args.space_z), mode="bilinear"),
           transforms.ScaleIntensityRanged(
               keys=["image"], a_min=args.a_min, a_max=args.a_max, b_min=args.b_min, b_max=args.b_max, clip=True
           ),
           transforms.ToTensord(keys=["image", "label"]),
       ]
   )

I am just wondering why in the testing phase, the Orientationd and CropForegroundd are not included?

Thanks！

[Kenton831]

Hi did you get any clues about that? I have the same confusion, thanks for replying