Good evening, I am trying to reproduce the results from Wang et al. using their cascaded convoloutional network with NiftyNet. Nevertheless when trying to train the CNN with :
net_run train -c train_whole_tumor_sagittal.ini --app brats_segmentation.BRATSApp --name anisotropic_nets.wt_net.WTNet
It indeed starts the brats_segmentation APP and creates the csv but I then get the following error :
Traceback (most recent call last):
File "/home/anaconda3/envs/NiftyNet/bin/net_run", line 11, in <module>
sys.exit(main())
File "/home/anaconda3/envs/NiftyNet/lib/python3.5/site-packages/niftynet/__init__.py", line 114, in main
app_driver.initialise_application(system_param, input_data_param)
File "/home/anaconda3/envs/NiftyNet/lib/python3.5/site-packages/niftynet/engine/application_driver.py", line 185, in initialise_application
data_param, app_param, data_partitioner)
TypeError: initialise_dataset_loader() takes from 1 to 3 positional arguments but 4 were given
which I don't know how to fix this ! So every help would be welcomed. Best regards
Paul
It seems net_run is calling an old version of the brats_segmentation app.
Using the latest model zoo version of the application should work, please see the readme: https://cmiclab.cs.ucl.ac.uk/CMIC/NiftyNetExampleServer/blob/master/anisotropic_nets_brats_challenge_model_zoo.md