DIAGNijmegen/pathology-hooknet-tls

Weights + input mask ?

ikramabdel opened this issue · 3 comments

ikramabdel commented

Great work ! thank you so much for sharing.
I was wondering if you plan to make the weights of your model public so we can test the inference code ?
and also if you can share how to obtain the input mask (the tumor segmentation algorithm is no longer available) so we can also test on another dataset.

martvanrijthoven commented

Dear AbdelIka,

Thank you for your interest in our work!

You can download the weights of the model here: https://zenodo.org/records/10614942

For input masks, a tissue-versus-background masks needs to be created. For new datasets you can use https://grand-challenge.org/algorithms/tissue-background-segmenation/. This should still be available, but please let me know if you have any problems with it and we can contact the creator.

Best wishes,
Mart

ikramabdel commented

Thank you !

Yes i have requested access on Tissue background segmentation and awaiting on a response.

But to clarify both inputs should be pyramidal images ?

martvanrijthoven commented

Yes you are correct both should be pyramidal.

Best wishes,
Mart

  • 👍1