CAMP-TUM network contributions: Quicknat and Daf3D

[vgonzalezd]

Fixes #5921

Description

@Al3xand1a and @ge96lip implement DAF3D[1] and Quicknat[2] networks and test them in open-source and local datasets. We use as a baseline the pytorch codes available in [3] and [4]
We are quite confident about the implementation, but feel free to contact us if you find errors. For any questions send us an email to ge45qix@mytum.de

We have some questions for the contribution:
For Quicknat:

1. we add the sequential class file in the networks folder because we do not know where to add it.
2. How do we include the squeeze and excitation requirement (sse and Csse) if it comes from a GitHub repository?
   https://github.com/ai-med/nn-common-modules/releases/download/v1.1/nn_common_modules-1.3-py3-none-any.whl
   For Daf3D
3. Are the overwritten blocks fine as they are? or do they have to be more flexible?
   For both:
   We run this command line (./runtests.sh --quick --unittests --disttests) but the error we are getting is not related to our changes, so we run our unit_test independently and they work. Same with documentation.

[1] Deep Attentive Features for Prostate Segmentation in 3D Transrectal Ultrasound.
Yi Wang, Haoran Dou, Xiaowei Hu, Lei Zhu, Xin Yang, Ming Xu, Jing Qin, Pheng-Ann Heng, Tianfu Wang, and Dong Ni.
IEEE Transactions on Medical Imaging, 2019.
[2] Roy, A. G., Conjeti, S., Navab, N., Wachinger, C., & Alzheimer's Disease Neuroimaging Initiative. (2019). QuickNAT: A fully convolutional network for quick and accurate segmentation of neuroanatomy. NeuroImage, 186, 713-727.
[3] https://github.com/ai-med/quickNAT_pytorch
[4] https://github.com/wulalago/DAF3D

Types of changes

• Non-breaking change (fix or new feature that would not break existing functionality).
• Breaking change (fix or new feature that would cause existing functionality to change).
• New tests added to cover the changes.
• Integration tests passed locally by running ./runtests.sh -f -u --net --coverage.
• [] Quick tests passed locally by running ./runtests.sh --quick --unittests --disttests.
• In-line docstrings updated.
• Documentation updated, tested make html command in the docs/ folder.

[wyli]

/black
It seems all the code reviews are addressed, thank you! there are two remaining test errors:

• mypy type checks https://github.com/Project-MONAI/MONAI/actions/runs/5333425006/jobs/9663876144?pr=6306 I think you can replicate it locally (linux bash) with

python -m pip install -U pip
python -m pip install -U -r requirements-dev.txt
./runtests.sh --mypy

please feel free to use # type: ignore if you think some lines are false alarms.

• torchscript test https://github.com/Project-MONAI/MONAI/actions/runs/5333425006/jobs/9663875599?pr=6306

======================================================================
ERROR: test_script (tests.test_daf3d.TestDAF3D)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/MONAI/MONAI/tests/test_daf3d.py", line 56, in test_script
    test_script_save(net, test_data)
  File "/home/runner/work/MONAI/MONAI/tests/utils.py", line 731, in test_script_save
    convert_to_torchscript(
  File "/home/runner/work/MONAI/MONAI/monai/networks/utils.py", line 746, in convert_to_torchscript
    script_module = torch.jit.script(model, **kwargs)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_script.py", line 1286, in script
    return torch.jit._recursive.create_script_module(
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 476, in create_script_module
    return create_script_module_impl(nn_module, concrete_type, stubs_fn)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
    create_methods_and_properties_from_stubs(concrete_type, method_stubs, property_stubs)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 393, in create_methods_and_properties_from_stubs
    concrete_type._create_methods_and_properties(property_defs, property_rcbs, method_defs, method_rcbs, method_defaults)
RuntimeError: 
Unexpected starred expansion. File a bug report:
  File "/home/runner/work/MONAI/MONAI/monai/networks/nets/daf3d.py", line 547
    
        attentive_features_maps = [self.attention(slf, mlf) for slf in single_layer_features]
        att_features, att_maps = zip(*attentive_features_maps)
                                     ~~~~~~~~~~~~~~~~~~~~~~~~ <--- HERE
    
        # second 4 supervised signals (af 1 - 4)


----------------------------------------------------------------------

I think the PR is ready to merge when these are fixed, please let us know if help needed cc @ericspod @Nic-Ma

[Al3xand1a]

It seems all the code reviews are addressed, thank you! there are two remaining test errors:

• mypy type checks https://github.com/Project-MONAI/MONAI/actions/runs/5333425006/jobs/9663876144?pr=6306 I think you can replicate it locally (linux bash) with

python -m pip install -U pip
python -m pip install -U -r requirements-dev.txt
./runtests.sh --mypy

please feel free to use # type: ignore if you think some lines are false alarms.

• torchscript test https://github.com/Project-MONAI/MONAI/actions/runs/5333425006/jobs/9663875599?pr=6306

======================================================================
ERROR: test_script (tests.test_daf3d.TestDAF3D)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "/home/runner/work/MONAI/MONAI/tests/test_daf3d.py", line 56, in test_script
    test_script_save(net, test_data)
  File "/home/runner/work/MONAI/MONAI/tests/utils.py", line 731, in test_script_save
    convert_to_torchscript(
  File "/home/runner/work/MONAI/MONAI/monai/networks/utils.py", line 746, in convert_to_torchscript
    script_module = torch.jit.script(model, **kwargs)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_script.py", line 1286, in script
    return torch.jit._recursive.create_script_module(
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 476, in create_script_module
    return create_script_module_impl(nn_module, concrete_type, stubs_fn)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 542, in create_script_module_impl
    create_methods_and_properties_from_stubs(concrete_type, method_stubs, property_stubs)
  File "/opt/hostedtoolcache/Python/3.8.17/x64/lib/python3.8/site-packages/torch/jit/_recursive.py", line 393, in create_methods_and_properties_from_stubs
    concrete_type._create_methods_and_properties(property_defs, property_rcbs, method_defs, method_rcbs, method_defaults)
RuntimeError: 
Unexpected starred expansion. File a bug report:
  File "/home/runner/work/MONAI/MONAI/monai/networks/nets/daf3d.py", line 547
    
        attentive_features_maps = [self.attention(slf, mlf) for slf in single_layer_features]
        att_features, att_maps = zip(*attentive_features_maps)
                                     ~~~~~~~~~~~~~~~~~~~~~~~~ <--- HERE
    
        # second 4 supervised signals (af 1 - 4)


----------------------------------------------------------------------

I think the PR is ready to merge when these are fixed, please let us know if help needed cc @ericspod @Nic-Ma

I fear I don't know what the torchscript bug is about. Why can't I use list expansion here? I appreciate any help!

[wyli]

/build

[wyli]

thank you! merging this if the test pipelines work fine