-
Notifications
You must be signed in to change notification settings - Fork 16
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
Multi-class tile import? #6
Comments
I've just created a tutorial on how to do multiclass import on the WSI-level here: #10 In some rare cases one would ideally also like to be able to import multiclass predictions from DeepMIB from exported patches from QuPath, but the current WSI-prediction approach will suffice in most cases I think. |
@SahPet Do you have a working groovy script that we could use as a drop-in replacement for the current script: Alternatively, we could offer it as an alternative method (for the multi-class case). |
Import of downsampled full WSI tifs is working well with multiclass, but I've struggled with adapting the tile script to multiclass (I think I managed once, but I can't find the version of the script). For now, I recommend using the multiclass import for full downsampled versions of the WSI: |
@SahPet That you for the info! OK, but that means that we still do not have a working multi-class tile import script, right? |
The current tileImport script does not support multiclass tiles, that is produced prediction tiles generated from MIB by a trained multi-class model.
It is possible to import predictions from MIB to QuPath by predicting on the full WSI in MIB which produces a stitched prediction image, which can be imported in QuPath using the importStitchedTIFfromMIBscript, similarly as done for FastPathology. And therefore, for the full pipeline a tile importer is not critical for using our pipeline.
However, there are scenarios where having a tile importer is beneficial over the other alternative, for instance if one only wishes to run prediction on a smaller region of the WSI (based on segmentations in QuPath) - which can be especially helpful if the image is extremely large. This can make inference a lot faster.
It is possible to do such a workflow in FastPathology, given that you provide the region of interest to run inference on. However, no such method has been made easily accessible. It would also require annotations or a model that can produce these ROI segmentation to use as a preprocessing step for the next patch-wise model.
Therefore, having multi-class support for the tile importer could be a valuable feature.
The text was updated successfully, but these errors were encountered: