HE mask reversion

[Johnny3333]

Thanks for your excellent work on Spatial Transcriptomics!

When I'm running spaHDmap demo on my own data, running:

`# Load the data from the 10X Visium folder
mouse_posterior = hdmap.prepare_stdata(adata=adata,
section_name='mouse_posterior',
image_path=image_path,
scale_rate=1)

hdmap.select_svgs(mouse_posterior, n_top_genes=3000, method='moran')
mouse_posterior.show()`
the mask was not on the correct position, indentifying the blank area as the H&E regions

[sldyns]

Hi @Johnny3333

It seems that the data may have been identified as Immunofluorescence. You can verify this by checking the image type of this STData object through the print function.

If that is the case, you can resolve the issue by specifying image_type as HE in the prepare_stdata function.

[Johnny3333]

I check the function and it shows the image type is HE, or should I modify the original code to correct it?

[sldyns]

@Johnny3333

Apologies for the inconvenience.

The current mask strategy was tuned based on our test data, so it may not be fully compatible with your specific dataset. We plan to implement a more robust mask strategy in future updates.

Regarding your current issue, I suggest two potential workarounds:

1. Set create_mask=False in prepare_stdata to skip the default masking.
2. Manually create a custom mask (or modify the source code logic), assign it to mouse_posterior.mask, and then proceed with the subsequent steps.

If you choose the second approach and find a solution that works well, would you mind sharing your strategy for creating the mask? It would be extremely helpful for us to improve the tool.

Thanks for your feedback!

[Johnny3333]

Thanks for your reply! It really helps me a lot. I'll try to figure it out as soon as possible~