Skipping converted slide when the file was not previously converted

[NasheethaMohm]

I ran into a problem when I was going through the conversion tutorial for macOS. It gave me this in the terminal when running the vsi2tif -i ./OS-2.vsi -o ./OS-2.tif -b bftools/bfconvert -v 0 command:

INFO:root:Performing single conversion...
INFO:root:Skipping already converted slide: ./OS-2.tif
Processing took 0.001 seconds

I have not converted this slide previously, so how can I solve this issue.

[andreped]

@NasheethaMohm There probably already lies a file with the exact same name OS-2.tif at the location where you are running the CLI from or where you are pointing to.

Perhaps a file was created when you first ran the CLI, had an issue causing the CLI to break. This may have created an "empty" file in the process.

If such a file exist, you can delete it, and then re-run the CLI.

[andreped]

@NasheethaMohm Did it solve the issue?

I could add some kind of overwrite flag, which ignores whether a file already exists, and overwrites it, if that makes your life easier?

[marcoaaz]

@NasheethaMohm There probably already lies a file with the exact same name OS-2.tif at the location where you are running the CLI from or where you are pointing to.

Perhaps a file was created when you first ran the CLI, had an issue causing the CLI to break. This may have created an "empty" file in the process.

If such a file exist, you can delete it, and then re-run the CLI.

I have the issue and it was not solved by your recommendation. Please, update me into what to do. I have seen the flag --noskip-converted does not have any effect.

My command:
"vsi2tif -i "E:/Feb-March_2024_zircon imaging/zircon_proj_VS200/Export/CA24MR-1_second_top.vsi" -o "E:/Feb-March_2024_zircon imaging/zircon_proj_VS200/Test/CA24MR-1_second_top2.tif" -b "C:/Users/acevedoz/OneDrive - Queensland University of Technology/Desktop/prices_software_vs200/bftools/bfconvert" --noskip-converted"

Thank you. 👍

[ziyuanzhao2000]

Same issue here!

[Filip9707]

I have the same issue, I am unable to convert my .vsi file to .tif.

Without the --noskip-converted flag I get the output:

INFO:root:Skipping already converted slide.

With the --noskip-converted flag I get the output:

INFO:root:Performing single conversion...
Processing took 0.000 seconds

Here is what I write in the terminal:

vsi2tif -i ./smbdata/dermatos_Filip/ExternalValidation/final_images/D16.978.vsi -o ./smbdata/dermatos_Filip/ExternalValidation/final_images/D16.978.tif -b /workspace/bftools/bfconvert

I followed the example notebook for the ubuntu/linux when installing everything.

[marcoaaz]

Dear friends, @ziyuanzhao2000 and @Filip9707

I am trying another library called python-bioformats followed by pyvips to save the VSI file as pyramidal OME-TIFF. My script still is in progress as I need to figure out how to write the XML description for it to be recognised as OME-TIFF in QuPath.

#region Dependencies

import os
vipsbin = r'c:\vips-dev-8.16\bin'
add_dll_dir = getattr(os, 'add_dll_directory', None)
if callable(add_dll_dir):
    add_dll_dir(vipsbin)
else:
    os.environ['PATH'] = os.pathsep.join((vipsbin, os.environ['PATH']))

import sys
import pyvips
import glob
print("vips version: " + str(pyvips.version(0))+"."+str(pyvips.version(1))+"."+str(pyvips.version(2)))

import javabridge
import bioformats
import matplotlib.pyplot as plt

#relative paths
from helperFunctions.mkdir2 import mkdir1, mkdir2 

#endregion

#region User input

image_path = r"E:\Feb-March_2024_zircon imaging\zircon_proj_VS200\Export\CA24MR-1_second_top.vsi"
sel_pyramid = 3 #sequential number of image #pyramid 3>7K pixels 
#sel_image = 2 #pythonic image count 
pixel_type = u'uint8'

#endregion

#region BIOFORMATS

javabridge.start_vm(class_path=bioformats.JARS,
                    run_headless=True)


dirname1 = os.path.dirname(image_path)
basename1 = os.path.basename(image_path)
folder1 = os.path.join(dirname1, "bf_" + basename1.replace(".vsi", ""))
mkdir2(folder1)

#Read metadata
xml1 = bioformats.get_omexml_metadata(image_path, url=None)      
xml2 = bioformats.OMEXML(xml1)

#General
acquisition_date = xml2.image(0).AcquisitionDate
layer_name = xml2.image(0).Name

#Specific
pixels = xml2.image(sel_pyramid).Pixels
x_count = pixels.SizeX
y_count = pixels.SizeY
c_count = pixels.SizeC
z_count = pixels.SizeZ
t_count = pixels.SizeT

path_list = []
for sel_image in range(0, 3):

    #output
    mode_str = f"{sel_image}"
    output_path = os.path.join(folder1, layer_name + "_" + mode_str + ".tif")

    image, scale = bioformats.load_image(image_path, c= None, 
                                         z= sel_image, t=0, series= sel_pyramid, index= None,
                                         rescale=False, wants_max_intensity=True, channel_names=None)
   
    bioformats.write_image(output_path, image, pixel_type, 
                        c=0, z=0, t=0, 
                        size_c=3, size_z=3, size_t=1, channel_names=None)

    path_list.append(output_path)

javabridge.kill_vm()

#endregion

#region PYVIPS

folder2 = os.path.join(folder1.replace("bf_", "vips_"))
mkdir2(folder2)

channel_list = []    
for file_input in path_list:  
    
    basename_output = os.path.basename(file_input)
    file_output = os.path.join(folder2, basename_output.replace(".tif", ".ome.tif"))

    im = pyvips.Image.new_from_file(file_input)    
    
    # openslide will add an alpha ... drop it
    if im.hasalpha():
        im = im[:-1]
    
    image_width = im.width
    image_height = im.height    

    im = pyvips.Image.arrayjoin(im.bandsplit(), across=1)
    
    im = im.copy()
    im.set_type(pyvips.GValue.gint_type, "page-height", image_height)
    im.set_type(pyvips.GValue.gstr_type, "image-description",
    f"""<?xml version="1.0" encoding="UTF-8"?>
    <OME Creator="OME Bio-Formats 7.0.1" 
        UUID="urn:uuid:b11ead14-703a-4817-a431-c0d32d795225" 
        xmlns="http://www.openmicroscopy.org/Schemas/OME/2016-06"
        xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
        xsi:schemaLocation="http://www.openmicroscopy.org/Schemas/OME/2016-06 http://www.openmicroscopy.org/Schemas/OME/2016-06/ome.xsd">
        <Image ID="Image:0">
            <!-- Minimum required fields about image dimensions -->
            <Pixels BigEndian="true"
                    DimensionOrder="XYCZT"
                    ID="Pixels:0"
                    Interleaved="false"
                    SignificantBits="8"                    
                    SizeC="3"
                    SizeT="1"
                    SizeX="{image_width}"
                    SizeY="{image_height}"
                    SizeZ="1"
                    Type="uint8">
                    <Channel ID="Channel:0:0" SamplesPerPixel="1">
                        <LightPath/>
                    </Channel>
                    <TiffData IFD="0" FirstC="0" FirstZ="0" FirstT="0" PlaneCount="1">
                        <UUID FileName="RL.ome.tif">urn:uuid:b11ead14-703a-4817-a431-c0d32d795225</UUID>
                    </TiffData>
                    <TiffData IFD="1" FirstC="1" FirstZ="0" FirstT="0" PlaneCount="1">
                        <UUID FileName="RL.ome.tif">urn:uuid:b11ead14-703a-4817-a431-c0d32d795225</UUID>
                    </TiffData>
                    <TiffData IFD="2" FirstC="2" FirstZ="0" FirstT="0" PlaneCount="1">
                        <UUID FileName="RL.ome.tif">urn:uuid:b11ead14-703a-4817-a431-c0d32d795225</UUID>
                    </TiffData>
            </Pixels>            
        </Image>
    </OME>""")

    im.tiffsave(file_output, compression="jpeg", tile=True,
                tile_width=512, tile_height=512,
                pyramid=True, subifd=True) #, bigtiff=True
    #none, jpeg, lzw, zip, and deflate

#endregion

I live in Australia, so this will kick-start your day if you are stuck with vsi2tif and what to try something else. I followed this documentation:

#https://pypi.org/project/pyvips/
#https://pythonhosted.org/python-bioformats/

CZI with Bioformats
https://forum.image.sc/t/python-bioformats-not-able-to-correctly-open-an-image/96600/12

My issue with pyvips and QuPath
libvips/libvips#2446

https://forum.image.sc/t/qupath-bioformats-high-compression-efficiency/58982/3

Tifffile (engine of python-bioformats) cannot write OME TIFF SubIDFs
https://forum.image.sc/t/python-bioformats-write-image-6d-images-series-handling/28633/7

Write pyramidal OME TIFF from pyvips: Here’s a slightly updated conversion script for libvips 8.14:
https://forum.image.sc/t/writing-qupath-bio-formats-compatible-pyramidal-image-with-libvips/51223/13

Alternatives to python-bioformats
https://forum.image.sc/t/ongoing-support-for-python-bioformats/49683/14

issue in ImageJ console:
Using TiffReader to determine the number of planes.

Writing TiffData (that is missing in the pyvips solution)
https://forum.image.sc/t/setting-up-ome-xml-for-a-new-microscope-from-scratch/62116/15
https://github.com/hzwirnmann/tcf_to_ometiff/blob/main/tcf_to_ometiff/tcf_to_ometiff.py
https://docs.openmicroscopy.org/ome-model/6.2.2/ome-tiff/specification.html#the-tiffdata-element

Have a nice day.
Marco

[adelinegadz]

Hi ! I have the same issue (MacOS here). Did anyone managed to solve it?
Thanks!

[andreped]

Hello, @NasheethaMohm, @marcoaaz, @ziyuanzhao2000, @Filip9707, and @adelinegadz!

Sorry for the slow response. I don't really work with digital pathology anymore, so I have not had time to maintain this project, but I wil try to address any issues today.

I am running some basic tests today. Are there any of you who have managed to get it working yet?

[andreped]

Managed to reproduce the bug:
https://colab.research.google.com/gist/andreped/0e945c30ebb01b309a36162d9ec133ae/vsi2tif-converting-olympus-wsi-to-generic-tiff-ubuntu-linux.ipynb#scrollTo=jDRNu7LAS2y2

Funny enough, it worked fine in Actions, but only because it did not raise an exception.

[andreped]

So I found that the latest patch contains a bug. If you use an older release it works just fine. So as a temporary fix, please just install an older release, like this one:

pip install vsi2tif==0.1.5

See my updated jupyter notebook here.

I will try to fix the bug and push a new patch very soon. So DO NOT use 0.1.6.

[adelinegadz]

Thank you so much for your quick response! It worked with the older version. Will try with my own slides now.