image.ndpi; too large??

[ahtiain...@gmail.com]

Hi,

I´ve tried to count positive cells from Hamamatsu images (ndpi) scanned with 40X, and with the largest ones (> 4 GB) it doesn´t work. I have set the maximum memory for 11 GB. Is there anything I can do, or should I re-scan everything with 20X to get them smaller?

Thanks,

Maarit

[micros...@gmail.com]

What specifically does not work?  The import process?  Does the cell generation start and then crash?

Also, if the image type says OpenSlide, there is a comment on the OpenSlide site about not knowing whether directories over 4GB will work, as 4GB is a file size limit for the TIFFs which ndpi files are built off of.  I do not have any actual experience with this file type :)

See here:

http://openslide.org/formats/hamamatsu/

[ahtiain...@gmail.com]

Exactly, the import process becomes stuck with the largest images :(

-Maarit

[Pete]

Hi,

This sounds like the issue described here: https://github.com/openslide/openslide/issues/174

By default QuPath uses OpenSlide for NDPI files.  The link above suggests that Bio-Formats can handle the larger NDPI images, so you could try installing the QuPath Bio-Formats extension from https://github.com/qupath/qupath-bioformats-extension

I think that QuPath will use Bio-Formats for all NDPI files if available, so hopefully this is enough to solve the problem.

Thanks,

Pete

[ahtiain...@gmail.com]

Many thanks,

 Bio-Formats extension solved the problem.

-Maarit