BraTS preprocessor hangs at t1c ants registration step

[KSchmidtACR]

Hello,

I've run the 0_preprocessing_single.py file from the BraTS repo on the example data with out issue. But when I change out the data files for in-house data, the processor hangs at the t1c ants registration step:

docker started!
connection established! sid: a26ea4fa85cc41868a6b823148777c0b
sid: a26ea4fa85cc41868a6b823148777c0b
status reveived:  {'code': 201, 'message': 'clientID json generation queued!'}
status reveived:  {'code': 202, 'message': 'client ID json generation finished!'}
sending input inspection request!
status reveived:  {'code': 201, 'message': 'input inspection queued!'}
status reveived:  {'code': 201, 'message': 'nifti examination queued!'}
status reveived:  {'code': 201, 'message': 'nifti parameters successfully sent to elephant client'}
status reveived:  {'code': 202, 'message': 'input inspection finished!', 'data': ['Follow Up_20230426']}
input inspection found the following exams:  ['Follow Up_20230426']
sending processing request!
status reveived:  {'code': 201, 'message': 'scheduling BRATS preprocessing!'}
image processing status reveived:
Follow Up_20230426 :  scheduling
status reveived:  {'code': 201, 'message': 'activating mode: cpu'}
status reveived:  {'code': 201, 'message': 'RAM detected: 59.834449768066406'}
status reveived:  {'code': 201, 'message': 'activating mode: auto cpu hd-bet'}
image processing status reveived:
Follow Up_20230426 :  enqueued
image processing status reveived:
Follow Up_20230426 :  copying raw files
image processing status reveived:
Follow Up_20230426 :  processing extra data
image processing status reveived:
Follow Up_20230426 :  t1c ants registration

Any insight into what is going wrong would be greatly appreciated. Thanks.

[neuronflow]

Thanks for your interest in BTK.

Hmm there must be a significant difference between your in house data and the example data. What is the shape of the data?

If you want to you can email me a sample and I can have a look.

[KSchmidtACR]

The size of the data are: t1 size: (512, 512, 182) t1c size: (512, 512, 182) t2 size: (512, 512, 61) flair size: (512, 512, 30)

Does this give you any ideas? Unfortunately I can't share the data with you. Can you share the location in the preprocessing container where I can look to troubleshoot this issue?

[neuronflow]

Hmm, that should not cause the problem here. I have an alternative preprocessing pipeline that runs niftyreg under the hood that you could beta test. If you are interested we can call.

Probably better to proceed this way, as it is intended to replace the preprocessing module from BTK.

[KSchmidtACR]

Sure, I can beta test. BraTS preprocessing is an important part of a workflow I'm implementing so whatever you think would best move things forward.

[KSchmidtACR]

Hello, would you be able to provide some more info this week on the new preprocessing pipeline you mentioned?

[neuronflow]

Hi Kendall, I contacted you on LinkedIn as I don't have an email address :)

[neuronflow]

The issue was caused by a dcm->nii conversion that went wrong. It would be interesting to have some of these files for debugging/development.