Importing a .vsi composite-image in ImageJ via bioformats-plugin - how do I get a high resolution Macro image

Hi there,
I have photographed microscopic pictures of brain slides using an overview function of my Olympus microscope. The resulting .vsi-file contains a picture which is stitched together of multiple single-shots. This I can open with the olympus software.
When importing it with the bioformat-plugin (since .vsi is not supported in imageJ by default), I get to select multiple series, only one of which includes the whole image (Macro-Picture). The resulting picture then is only of very poor quality that can’t be further used for cell counting. However other series give me very high resolution of a random picture-area.
So how can I import the whole picture in high resolution? Thanks so much, I will send Cake-Emojiis to anyone who can help :wink:

@TMM-98

Have you tried BigDataViewer yet? Might be a good solution for you in this case…

Loading entirely the full resolution of your image is not an option because the size of the uncropped plane will unlikely exceed the 2GB limit.

Most viewers dealing with this type of imaging modality usually access the data by navigating across different resolution levels as well as by loading each resolution in tiles as requested.

The default ImageJ viewer historically used by the Bio-Formats plugin was never designed for this type of data. For multi-resolution images, either QuPath which was developed specifically with whole slide imaging in mind or the BigDataViewer plugin already mentioned by @etadobson should be suitable options.

1 Like

Thank you for your answers!

Unfortunately, QuPath refuses to open my .vsi-files regardless of all my attempts.
The Big Data Viewer seemed promising as well, but gives a black picture as an output as well.
I have attached a video with my file-opening workflow - probably I am just not smart enough to use the plugin right. Would you or someone else mind taking a look on what my problem is exactly, since I don’t find any specific information in the web…?
PS: I have tried to select more series in the big data viewer simultaneously, but that gives pretty much the same results.

If you are interested in getting it to work though QuPath, please specify what you’ve tried. VSI files supposed by Bio-Formats should work with the latest milestone released (no further steps needed), or v0.1.2 (if you’ve installed the Bio-Formats extension).

If it still fails, reporting any specific error messages or sharing a file would be needed to understand what is going wrong.

See also https://github.com/ome/bioformats/issues/3533. Is this a duplicate of this issue or a separate issue @TMM-98? In the first case, let us agree on one place for support to help keeping the discussion focused.

@s.besson : I only wanted to separate the fiji from the qupath approach. But yes, let’s keep it here - Thanks!

@petebankhead : I now have opened my file in qupath. But still my problem is that only a fraction of the whole slide is displayed: As attached, the QuPath-Picture only represents a small part of the Preview Picture.
ImageJ does the same - and I previously thought that this was only due to the size of the picture… but Qupath should be able to deal with that, right?

Yes, the size shouldn’t be an issue for QuPath here.

Although upon re-reading your first post, I see you describe “I have photographed microscopic pictures of brain slides using an overview function of my Olympus microscope” - the .vsi images I’ve seen have come from an Olympus slide scanner, so I’m not sure how they might differ.

Can you share a screenshot of how the exact same image appears within the Olympus software? And ideally share a link to an original .vsi file…?

1 Like