Error using the trained U-Net model

[RuiqiGeng]

Hi, I am trying to use the pre-trained U-Net for my own CT data to output airway segmentation in directory "testdata/" but ran into these errors:

In FILE '/gstore/home/gengr3/BronchiNet/bronchinet/bronchinet-master/src/common/workdirmanager.py' and LINE '36': ERROR: File '/gstore/home/gengr3/BronchiNet/bronchinet/testdata//referenceKeys_posteriors.npy', does not exist... EXIT

In FILE '/gstore/home/gengr3/BronchiNet/bronchinet/bronchinet-master/src/common/functionutil.py' and LINE '185': ERROR: No files found in '/gstore/home/gengr3/BronchiNet/bronchinet/testdata//Posteriors/' with '*'... EXIT

In FILE '/gstore/home/gengr3/BronchiNet/bronchinet/bronchinet-master/src/common/functionutil.py' and LINE '185': ERROR: No files found in 'testdata//BinaryMasks/' with '*'... EXIT

What kind of data are needed in referenceKeys_posteriors.npy, Posteriors/, and BinaryMasks/?

Thanks!

[id-b3]

Hello,

Can you please add more information about how you are launching the experiment? Are you using the docker image or a locally downloaded repository? What file format are your scans in?

[RuiqiGeng]

Hi, Thanks for your response! I am using a locally downloaded repo and the CT scans are in nii.gz. This is my tree structure:

workDir is my work directory. The bash command I used to run the pretrained model is: bash models/run_model_trained.sh BaseData/ testdata/ --torch$

[Nandayang]

Same issue, anyone solve that?

[dscarmo]

Is this because the models are on git lfs maybe?

[m5b6]

Same issue here