Using scyjava to convert resolutions pyramid VSI file to tif file resaults with 1/40 resolution layer of the pyramid only.

[OriKovacsiKatz]

Hello
My name is Ori , I am an HPC and AI advisor at the Technion (Research Institute and University)
I advise a researcher and need help with medical imaging Olympus resolution issue:
(continuing my email with Edward Evans suggesting me opening my inquiry to the public eye)
Using imageJ with scyjava to convert VSI (Olympus) images to tif format , provide only a small portion
of the imaging binary data.
Specifically there is some imaging images pyramid where the x40 resolution required so the resulting
image is only X 1/40 of the necessary binary details.

How can I force bio-format convert to export tif image with the x40 zoom size out of the VSI Olympus pyramid?

I use the code:

import imagej
from os import listdir 
import numpy as np
import os.path
ij = imagej.init(headless=False)

SOURCE_DIRECTORY=r"/home/<USER NAME>/Data/Melanoma_WSI/H&E/20241027/"
TARGET_DIRECTORY=r"/home/<USER NAME>/Data/Melanoma_WSI/H&E/20241027_converted/"
def convertImageToTif(image_name): 
    image_url=SOURCE_DIRECTORY+ image_name
    print("# read image:")
    print(image_url)
    save_url=TARGET_DIRECTORY+image_name.split('.vsi')[0]+'.tif'
    print("# write image:")
    print(image_url)
    jimage = ij.io().open(image_url)
    print("# after opening image...")
    ij.io().save(jimage, save_url)
    print("# after saving image....")

files = np.array(listdir(SOURCE_DIRECTORY))
# files in source which contain ".vsi" 
files = files[np.flatnonzero(np.core.defchararray.find(files,'.vsi')!=-1)] 
print("# converting files:")
print(files)


for file in files: 
    convertImageToTif(file)

[ctrueden]

@OriKovacsiKatz The best place to discuss this sort of question is https://forum.image.sc. You can post in the Development category with tags pyimagej and bio-formats. My advice would be to first get this working without PyImageJ in the mix; at its core it's really a question about Bio-Formats and/or SCIFIO.

Here are some quick thoughts:

• The ij.io().open(image_url) is a quick convenience method to open the first image in a file. But it won't open all images in the file; in the case of multires, Bio-Formats might be returning multiple 5D images.
• You could try using images = io.scif.img.IO.openAll(image_url) which should return a list, from which you can select the 40x one.
• Failing that, you could record a script (set Macro Recorder to JavaScript or Python) of yourself importing the correct image using Bio-Formats from the Fiji GUI, and then adapt the outputted script line into your program here.

I'm going to close this issue since it's not really something to be improved in the scyjava Python library, but I warmly encourage you to post your question on forum.image.sc if you are still stuck, tag with @ome and/or @ctrueden if you like, and hopefully we can help further!