Viewing 3D Prairie View stacks of .tif files

Hi! I am not really sure if BioFormats is intended to handle this situation so, first a description!

I have a Prairie View set of tif files along with their xml and env files, with all of the appropriate metadata included to rebuild a 3D volume (PVScan version=“5.3.64.400” , BioFormats 5.8.2 first, and then 6.0.0). The Z stacks are not all equal in size since this was scanned on a live sample, and it is most efficient to only scan the area of interest (I wish this could be done on our 880!). That means the final volume would not be a cuboid.

Should it be possible to put together and view this kind of image with BioFormats? Currently when I open the xml file with Fiji/BioFormats plugin, it reads each tile as a separate Z stack and opens each correctly (with “Stitch tiles” checked). It does not put them all together though, in terms of their relative X Y and Z coordinates. When using other software and attempting to open these images, I generally see one of these stacks (QuPath (via BioFormats), Imaris).

Imaris reads each stack as a time series (not sure if that is relevant as I think they have their own reader).

I have hosted a sample data set here: https://drive.google.com/drive/folders/1tqyhZij6q7hMp9w7S0JsEpWquTzWfzQK?usp=sharing

Please let me know if there is any other information I can provide.

Cheers, and thank you for your time.

Hi,
disclaimer: I have no idea what is Prairie View, but I guess it is not “a city in Waller County, Texas, United States”

Do you have z-stacks in time?

The Z stacks are not all equal in size since this was scanned on a live sample

What do you mean by that? Do you mean that your XY resolution is not constant between the stacks? Or your width/height/depth (in pixels) are different?
In either way you probably have to register the datasets in order to open them together and it may be complicated…

Ilya

Haha, Prairie Bruker format. https://docs.openmicroscopy.org/bio-formats/5.9.2/formats/prairie-tech-tiff.html

It is a single time point. The Z stacks all have the same XY resolution and size, but they start/stop at different heights.
*Sorry, for clarity, each set of XY frames is in the same plane, but there are different numbers of frames in each stack. So you could have a 2x2 all at height Z, and then a 3x3 at Z+1, but there is no stack with a Z+0.5.

right, I am quite naïve :wink:
It seems that I am quite tired today to understand or quite stupid or, most likely, both…
How about just to add a dummy tif image that is totally black, where you have a missing section?

I actually wonder about good answer to your problem. To me it sounds that you have to combine those images in a program that is capable of manipulating with matrices and shift those according to the coordinates that you have in metadata (you can do that in Matlab, for example). So, you can use BioFormats to read the data, but after that you have to write an own procedure to shift datasets according to their xyz coordinates and save in, for example HDF5 format.

Hmm, interesting. I’ll have to give the dummy tif file a test. Or when I get access to the microscope at least see if the issue persists with a basic 2x2x2 image set. I asked for a smaller example, but do not normally have access to that microscope. I will next week though…!

I have tested the sample fileset provided with the latest Bio-Formats and currently it will only open the series individually rather than fusing them which you would have to do afterwards. The way Bio-Formats determinds the number of series is by parsing the metadata XML file and each Sequence element will be created as its own series.

As a side note if when opening Prairie files you should open the XML or config file, if you just select the tif files it may be read by one of the Tiff readers. In this particular case it does not solve the issue however.

Ah, thanks for letting me know. I was using the XML file, but I won’t beat my head against this part if it isn’t going to work.

Cheers!