Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Whole slide dataset to extract random patches #4099

Open
drbeh opened this issue Apr 10, 2022 · 2 comments
Open

Whole slide dataset to extract random patches #4099

drbeh opened this issue Apr 10, 2022 · 2 comments
Labels
Pathology/Microscopy Digital Pathology and Microscopy related

Comments

@drbeh
Copy link
Member

drbeh commented Apr 10, 2022

Currently, PatchWSIDataset generates patches based on pre-defined locations. In digital pathology use cases, there is a need for generating such patches randomly from all over the whole slide image. This requirement is different than using TileOnGrid to generate such patches since this transform is operating on the image in the memory while RandPatchWSIDataset directly deal with whole slide images on the file system.

@vikashg
Copy link

vikashg commented Apr 13, 2022

Can we also build in the capability to extract patches at different zoom levels. That will be ideal.

@drbeh
Copy link
Member Author

drbeh commented Apr 13, 2022

@vikashg thanks for your input. We have recently addressed that in PR #4036, and now you should be able to pass level argument to extract patches at that level.

@drbeh drbeh moved this from Backlog to Todo in AI in Pathology🔬 Apr 25, 2022
@drbeh drbeh self-assigned this May 16, 2022
@drbeh drbeh moved this from Todo to In Progress in AI in Pathology🔬 May 19, 2022
@drbeh drbeh moved this from In Progress to Backlog in AI in Pathology🔬 May 27, 2022
@drbeh drbeh removed their assignment Jul 11, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
Pathology/Microscopy Digital Pathology and Microscopy related
Projects
Status: No status
Development

No branches or pull requests

2 participants