Efficiency benchmark compared to SpatialData based IO?

[rushin682]

Hi,

this looks like a great tool for WSI images IO. Thank you for your efforts on this!

I saw the issue on SpatialData regarding this, and I wonder, if there has been an read IO comparison between SpatialData and wsidata for the same?

We have a lot of Spatial-omics samples with corresponding full res WSI images and I wish to use your tool for multi-modal tiling (i.e IMAGE + Gene Exp/Protein Exp pairs) provided it's much more efficient compared to SpatialData.

Happy to provide a better Idea of what I intend to achieve and discuss further. Thanks once again for this tool :)

best,
Rushin

[Mr-Milk]

Hi, thanks for the interest in wsidata. wsidata is built on top of spatialdata. The additional feature is that it can open pyramids image formats that are not compatible with OME-ZARR. In terms of read speed of image, this depends largely on the underlying file format and the backend used for IO. Therefore, it's hard to have a fair and reasonable benchmark for spatialdata on IO.