Help on exporting pathology images and corresponding annotations

We loaded SVS (Aperio) scanner files using Qupath for annotation, and we’re now attempting to extract matching pathology and corresponding annotation images. In order to extract TIF from SVS, we used Aperio’s Digital Slide Studio, and was only able to extract an area of interest due to the size of each file. We are able to independently extract our annotation as images (below). However, because the two extractions were done independently, they don’t match in size, edge and dimension, and x,y-coordinates.

Could you help us with if there’s a way to register/match these two independently extracted images? Is there a way to do this with a one-step process using Qupath? Any idea is welcome.

Thank you very much,

Mike

There is information about the Interactive image alignment, both for manual use and through scripting, scattered about the forum. If you are aligning two or more images, that is usually the way to go. It is a rigid or affine transformation only, however, so how useful it will be depends on your type of analysis and sample preparation.

You can also create ROIs in the SVS files directly that could be used to export smaller chunks of the images if you want that.


The whole slide image could even be exported if needed, as an OME-TIFF.
https://qupath.readthedocs.io/en/latest/docs/advanced/exporting_images.html

You have only shown a single masked image, of what looks like the pixel classifier mask. I’m not sure how to use that to give any advice.

1 Like