Update wholeBody_ct_segmentation

[KumoLiu]

Description

While testing the lazy-resampling benchmark, I find that maybe there's room for improvement in the pre-processing.
In this PR, I mainly update several places below:

• Remove NormalizeIntensityd since zeros aren't the background in the CT
• Move ScaleIntensityd in front of the CropForeground and change the value to 0-1 since we can only pad zero now in the lazy mode.
• Remove one-hot in the AsDiscreted which may save memory
• Update the metric in the training part from Accuracy to Dice
• Using the original label in the validation due to this issue

Note: the config was only tested on low resolution

cc @tangy5 @wyli

Status

Work in progress

Please ensure all the checkboxes:

• Codeformat tests passed locally by running ./runtests.sh --codeformat.
• In-line docstrings updated.
• Update version and changelog in metadata.json if changing an existing bundle.
• Please ensure the naming rules in config files meet our requirements (please refer to: CONTRIBUTING.md).
• Ensure versions of packages such as monai, pytorch and numpy are correct in metadata.json.
• Descriptions should be consistent with the content, such as eval_metrics of the provided weights and TorchScript modules.
• Files larger than 25MB are excluded and replaced by providing download links in large_file.yml.
• Avoid using path that contains personal information within config files (such as use /home/your_name/ for "bundle_root").

[tangy5]

Thanks for the updates, looking forward to seeing the complete updates.

[Nic-Ma]

Hi @KumoLiu ,

Any update here?

Thanks.

[KumoLiu]

Hi @Nic-Ma, I don't have any update for this PR so far. The initial goal of this PR was to document what I thought could be improved in this model when I measured lazy resampling. However, because the data contained within this model zoo is actually high resolution, I was testing it on low resolution. So I'm just putting this PR here as a reference. If you think we need to keep updating it, I may need to redo some high-resolution experiments.
Thanks!