About writing a h5 file on BraTS validation dataset

[chengjianhong]

@ellisdg Hi, I think there is a insufficiency on writing a h5 file for BraTS validation dataset(without ground truth) in code. Because it needs a ground truth data together with four modalities data. I make it by copying a ground turth data into every subject path, which is unadvisable. So I think It needs to update the method to generate h5 files for the validation set or testing set.

[nysuka]

to generate h5 files without truth data for testing, you need to write out the h5 files without the truth storage, so in data.py, you can modify def create_data_file to skip the truth_storage with a flag or a separate function name and call it in write_data_to_file. during prediction, in prediction.py, modify run_validation_case by deleting all processing of truth data.

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions. If you have questions, feel free to join the Slack group or email me at davidgellis2@gmail.com.