[ome-devel] Trello Bio-Formats 5.6.0 board closed / ImageJ BioFormats Image Importer Plugin

[Michael Ellis]

Dear all,

ITAC?= Is This Assumption Correct?
IJBFIIP = ImageJ BioFormats Image Importer Plugin

I am at a loss, can you point me in the right direction?

Sometime back I submitted a pull request on the IJBFIIP so that image slice labels could be set from channel names stored within .OME.TIFF files.

I was informed this pull request would be shelved as it amounted to a functional change and not just a bug fix.

Eventually I was pointed at using Trello and to follow the  Bio-Formats 5.6.0 board. This I did and added some further comments in relation to the proposed final solution.

I sit, i wait, I soldier on with my own patch to the IJBFIIP.

Yesterday I see that the Trello Bio-Formats 5.6.0 board is closed and I do not know where to go to track what is happening to the proposed changes for the next major release?

Further, I am implementing a computer assisted microscope package (SmartCapture 4), I’ve chosen to use the BioFormats.jar package and save in OME.TIF format as this is a nigh on perfect match for our requirements of being able to save 5D images will instrument and experiment metadata - Yipee!

My intention is for SmartCapture 4 to hand off image processing to other applications, principally  ImageJ/Fiji and our own SmartType karyotyping application. Transferring images as OME.TIF files is ideally suited.

However, I do have further observations in respect of the IJBFIIP that I would at least like to put forward without feeling they these requests just disappear into the wind.

Specifically,

If the .OME.TIF file format is considered the most complete or canonical image file format for OME metadata (ITAC?) then it would seem at least desirable that the IJBFIIP utilise as much metadata from  .OME.TIF files as possible (for instance channel colour as well as channel name).

I do see that there is a potential for semantic mismatches or ambiguities in how some metadata might be used, but even if only a part of the metadata is used to set ImageJ ImagePlus image properties, should not *all* the metadata be accessible *after* the IJBFIIP has completed the import?

The argument that I can patch IJBFIIP and do whatever I want seems a weak one. If I have to patch the IJBFIIP, I am burdened with having to track and merge all of your changes (as will anyone else wanting something extra from the metadata).

My suggestion (is this already possible?), is that the IJBFIIP provides a mechanism for accessing the metadata for any ImageJ ImagePlus that has been imported by the IJBFIIP.

If (as I suspect) there is no hook on an ImagePlus for hanging the XML metadata, then perhaps this could be hived off into a file and reference to that file could be attached to the ImagePlus? I’d then at least be able to write my own ImageJ plugins that could inspect an ImagePlus, determine if it had associated OME data and use the BioFormats.jar library to gain access to the extra metadata without having to have my own version of the IJBFIIP.

I really want to make use of BioFormats, it’s such a perfect match for our needs. But, having IJBFIIP filter our so much of the metadata seems just counter intuitive if, as I suspect, ImageJ is the most widely used application for reading OME.TIF files?

With best regards to all — Michael Ellis
Digital Scientific UK Ltd