Add unet++ notebook

[nhhung1810]

Fixes #1475.

Description

A few sentences describing the changes proposed in this pull request.

Checks

• Avoid including large-size files in the PR.
• Clean up long text outputs from code cells in the notebook.
• For security purposes, please check the contents and remove any sensitive info such as user names and private key.
• Ensure (1) hyperlinks and markdown anchors are working (2) use relative paths for tutorial repo files (3) put figure and graphs in the ./figure folder
• Notebook runs automatically ./runner.sh -t <path to .ipynb file>

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[wyli]

thanks, please follow this https://github.com/Project-MONAI/tutorials/blob/main/CONTRIBUTING.md#create-a-notebook to create the necessary cells, I think you can copy them from https://github.com/Project-MONAI/tutorials/blob/main/modules/UNet_input_size_constrains.ipynb

[nhhung1810]

After several trial, I finally understand the basic process, hope you guy can check on the notebook content :D

[mingxin-zheng]

Thank you @nhhung1810 for the PR. I just started to read the notebook. I noticed that there are many commented code, and the last cell has outputs from all-commented code block. May I know what the commented code is for?

[nhhung1810]

Hey @mingxin-zheng , that commented code is for exploration on the implementation of UNET++. I'll delete it after the other things resolved

[wyli]

thanks! would be great to have a conclusion section to recommend the usage, for example, about how to set number of downsampling layers and input spatial sizes, and the normalisation argument might require a tuple for additional parameters https://github.com/Project-MONAI/MONAI/blob/6f5cea3795278c20568261ebb38907d94b576b98/monai/networks/nets/daf3d.py#L490-L491. the current notebook explores many tests but it's unclear to me what are the suitable configurations for this model.

[nhhung1810]

Sure @wyli, I'll dig into the original papers and others papers to write the conclusion. Also, I'll include a network image for reference

[nhhung1810]

Just add some conclusions to the normalization (also an example of passing norm's argument) and up/down-sampling sections. I also comment about the features argument as I found it have a small difference compare to the paper.

[wyli]

Thanks, it looks good to me.