Data import bug from Leica containing FLIM and Tilescan (from Navigator)

Hi there,
when trying to open a Leica LIF-Project, which contains an Tilescan and at least two images with FLIM data, from the second one, data are not read correctly.
Projects with Tilescan and data without FLIM work fine.
When Cropping the Tilescan (inside Leica software) data without FLIM can be red although there are some error messages.
When Cropping the Tilescan (inside Leica software) data with FLIM cause Exception.

Unfortunately I was not allowed to upload data, but you can download some datasets to play with here:
https://www.mh-hannover.de/cellneurophys/zeug/Bioformats/Leica_Dataset.zip

Please let me know if you need further information to handle the issue, or if you wish to have further examples.
Surely I used the latest version of FIJI & bioformats etc.

I would also very much appreciate to get access to additional Metadata from Leica files. Is there a way to support you in this respect?

Thanks in advance

Andre

PS: same issues accur when importing the files to Matlab (using bioformats.jar and bfopen.m), which is my final goal.

1 Like

@AndreZeug

The Bio-Formats team does not frequent this forum too often… so I would advise you to go to their website directly to report this issue. That will be the best way to get a fix.

UPDATE: See @joshmoore’s response below! And thank you for reporting here… Bio-Formats is a Community Partner! :slight_smile:

1 Like

Hi @etadobson and @AndreZeug.

Quick clarification. Since Welcome Bio-Formats as a community partner! the Bio-Formats team does track this tag. (In fact, @etadobson, the “submit bug” link under “report this issue” points backs here. :slight_smile: )

So :+1: for the question and thanks for the sample data. No need to repost anywhere else. The UK crowd however is now out of business hours so it might not be until the European morning for a response. (Of course, if anyone else has suggestions, feel free.)

All the best,
~Josh

3 Likes

Ha! Sorry @joshmoore !!! I had the ‘old’ response just too ready-to-go. :slight_smile:

1 Like

Hi @AndreZeug, thank you for providing a sample dataset we can test with. I was able to reproduce the issues you described with that dataset and the latest Bio-Formats.

For the FLIM images, are any of the dimensions or metadata incorrect in the OME-XML which is displayed? (Particulary pixel type)

Also you mentioned additional metadata which you dont have access to. Is there a list of fields or properties that you know are missing?

I have created a Bio-Formats Trello card to document the problems and track the issue: https://trello.com/c/CyA4IMrK/358-lif-with-flim-and-tilescan

Dear @dgault, many thanks for taking care of this issue.

FLIM-Images: I could not identify any errors in OME-XML, however the XML which one can receive from the Leica software contains many more details. To better see what I mean I prepared another snap
https://www.mh-hannover.de/cellneurophys/zeug/Bioformats/Leica_Dataset2.zip
I acquired a “Sequence” of 2 channels x 4 conditions, varying laser power. Leica handles sequences as additional channels which might be confusing, but is probably due to practical reason of dimensionality since sequences are allowed to have varying number of channels.
The Leica XML I obtained from saving “Image Properties”. So maybe Leica could provide a library to read all relevant information?
I also provided some screen shots from the Leica software that one can get an impression which parameters are displayed.

OME-XML: Would be nice to see basic imaging conditions (in this case including laser power) for all channels / sequences. Is probably too much to include all parameters.

FLIM-data: is there any chance to read FLIM-data using bioformats library? (Or any plan to include?) So far I use the Leica software to “Export” FLIM-data as ‘PTU’-file (PicoQuant Time Tagged File). This however does only contain the ‘TCSPC time’ (photon time in relation to previous laser pulse) but not the ‘time tag’ (from which laser pulse period the photon originates). So additional filtering or correlations could not be applied to those data.

Dear @dgault, dear all,
checking more carefully OME-XML metadata from Leica LIF-projects, I realised that OME-XML misses important information such that it is hardly possible to get a clue of the acquisition parameters. Here you can find an example including XML-metadata provided by Leica software:
https://www.mh-hannover.de/cellneurophys/zeug/Bioformats/Leica_Dataset_XML.zip

In OME-XML important information like ExcitationWavelength / Power and EmissionWavelengthMin/Max, DetectorGain, DetectorIsActive per channel are missing or assigned to the wrong channel.

Since this Metadata issue is not related to the topic of the tweet please feel free to move to a separate topic, if appropriate.

Thanks, A

Thanks @AndreZeug for providing those samples, I will go through them and make a list of any missing fields. It is worth pointing out that a lot of the fields that do not fit with the OME-Model will still be stored as global metadata, in FIJI this is displayed in the Show Info dialog rather than the OME-XML window.

Dear OME Team, is there already some progress on the topic(s) of data import from Leica containing FLIM and TileScan and the reading of MetaData? Maybe I missed it.
Please let me know if I can do something (test, provide extra data, etc.)

Interest to get this issue solved is still high :smiley:

Hi @AndreZeug, Im afraid we don’t have a fix for this issue yet. We just released Bio-Formats 6.1.0 today which has seen us focused on completing the latest phase of our work to support binary vessels as well updating a number of the dependency components. With that released we will look at integrating some format fixes for smaller, more frequent patch fixes. I will review and retest with the dataset you provided this week to see if I can resolve the issue and get a fix put in place.

1 Like