Update HoVerNet pipeline

[KumoLiu]

Fixes #.

Description

add training mode descriptions in README
update prepare_patches.py
change to "fast" mode

Checks

• Notebook runs automatically ./runner [-p <regex_pattern>]

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[bhashemian]

Hi @KumoLiu, it looks good to me. just few comments of the texts.

[bhashemian]

@KumoLiu, could you please update the default mode of other hovernet pipelines too? thanks

tutorials/pathology/hovernet/evaluation.py

Line 142 in 5a49733

parser.add_argument("--mode", type=str, default="original", help="choose either `original` or `fast`")

tutorials/pathology/hovernet/inference.py

Line 189 in 5a49733

parser.add_argument("--mode", type=str, default="original", help="HoVerNet mode (original/fast)")

[KumoLiu]

@KumoLiu, could you please update the default mode of other hovernet pipelines too? thanks

tutorials/pathology/hovernet/evaluation.py

Line 142 in 5a49733

parser.add_argument("--mode", type=str, default="original", help="choose either `original` or `fast`")

tutorials/pathology/hovernet/inference.py

Line 189 in 5a49733

parser.add_argument("--mode", type=str, default="original", help="HoVerNet mode (original/fast)")

Of course, have updated!

[bhashemian]

@KumoLiu this one needs to be fixed too. Thanks

tutorials/pathology/hovernet/inference.py

Line 136 in 5a49733

model.load_state_dict(torch.load(cfg["ckpt"], map_location=device)["net"])

[KumoLiu]

Hi @drbeh, I have already updated the performance curve, could you please double-check the license descriptions in the README, if that's ok I think this PR can be merged, thanks!

[wyli]

in the readme and notebookt I think we should cite this paper https://www.sciencedirect.com/science/article/abs/pii/S1361841519301045 instead of the arxiv version.

do we need an acknowledgement section for the original author/implementations? the detection folder's readme is a good example https://github.com/Project-MONAI/tutorials/tree/main/detection#acknowledgement cc @drbeh @JHancox

[bhashemian]

in the readme and notebookt I think we should cite this paper https://www.sciencedirect.com/science/article/abs/pii/S1361841519301045 instead of the arxiv version.

do we need an acknowledgement section for the original author/implementations? the detection folder's readme is a good example https://github.com/Project-MONAI/tutorials/tree/main/detection#acknowledgement cc @drbeh @JHancox

I agree with @wyli! although we don't have any external contributor to acknowledge.

@KumoLiu would it be possible to update them using the following citation? thanks

Simon Graham, Quoc Dang Vu, Shan E Ahmed Raza, Ayesha Azam, Yee Wah Tsang, Jin Tae Kwak, Nasir Rajpoot, Hover-Net: Simultaneous segmentation and classification of nuclei in multi-tissue histology images, Medical Image Analysis, 2019
https://doi.org/10.1016/j.media.2019.101563