Johannes Soltwedel
@jo-soltwedel
Research software engineer @gerbi-gmb.de for all things #OME-ZARR, #NGFF, FAIR data. Physicist, Python code-jockey, train enthusiast.
Not going to be in Cracow for the big event so had to make do with a small impromptu @napari.org x @openmicroscopy.org #omezarr Hackathon in Bern 🔬🇨🇭 @m-albert.bsky.social @guiwitz.bsky.social
Just passing through, but wouldn't miss this stunning piece of #brutalism - St. Joseph in #lehavre, France
...or image #stitching, where a microscope 🔬 could just write out multiscale tiles on-the-fly along with the coordinates of each tile in the image. An application like @napari.org can then just parse these in and forward it to tools like multiview-stitcher.github.io/napari-stitc...
Still, what's new? My personal favorite: The scene 🎥 Thanks to a now standardized set of transformations, you can write out the spatial relationships between (multi-modal) images, which is super neat in image registration... (Dataset: rodare.hzdr.de/record/3131)
First, why 0.6rc0? The door is closed for changes to the spec after we allowed ourselves some iterating (orange arrow). However: we need to give developers the time to update their implementations (green arrow) before the #NGFF RFC process allows us to hit the switch ⚡on a formal 0️⃣.6️⃣
That's one satisfying github notification list to end the day with 🏝️ Now just one more thing to do...
Great interactive talk about surveys/interactive survey about talks and cognitive biases from my colleague @lisa-schaefer.bsky.social from @gerbi-gmb.de at #ELMI2026
Just gave a talk at #ELMI2026 and a panel discussion about how #FAIR data impacts #AI and its use in #bio-image-analysis 🔬🧪- and what we at @gerbi-gmb.de do about it 👷👷♀️ Slides available here: zenodo.org/records/2070...
Just some #ngff enthusiasts out in #dresden @perlman.bsky.social
A painful truth well put by @virginieuhlmann.bsky.social at #OME2026 😬 Striving for software #interoperability will make all our lives easier in the future!
Finally turning a lot of Github handles into real-world acquaintances and meeting many friends from the field 😀 Looking forward to three days of @openmicroscopy.org #OME2026
I'll never get tired of lazily browsing elastix-coregistered #WSI data in @napari.org using the upcoming @openmicroscopy.org ome-zarr @zarr.dev format version that allows to specify and read transformations directly from the image data itself. Data source: rodare.hzdr.de/record/3131
Kind of flew under the radar, BUT we released napari-omero 0.6 recently 🥳 @openmicroscopy.org You can now down-/upload ROIs from and to OMERO inside @napari.org 🤩Bringing the power of napari for annotation to the management of OMERO :) Thanks @psobolewskiphd.bsky.social and other maintainers!
Ha! Just noticed that my terminal now (has it always?) shows a small loading bar 🚧 when @pytest.org tests are running! Cool :)
Hopping on the ✨ 🚅 to Dresden after a week of #OME_NGFF_Hackathon. Looking forward till next time 😃
This is what hacking *actually* looks like. #OME_NGFF_Hackathon
Whenever I feel down, I have a 👀 at the download statistics of that #opensource data-repository I and my team at OncoRay (Dresden) compiled during my PhD at @tudresden.bsky.social @hzdr.bsky.social. 😌 rodare.hzdr.de/record/3131
Suitable @poldresden.bsky.social farewell cake for a #bioimage analyst? 🥮🤤
Broken gif in original post 😱...better look at the new feature of the @napari.org-clusters-plotter here 👇
#bioimage 🔬 data #visualization and #exploration in @napari.org with the napari-clusters-plotter (0.10.0) becoming yet a bit more versatile: Select objects & put them into new layers for processing, introspection or whatever. Works for all layers, of course! #bioimageanalysis @mazoc.bsky.social
@stadt-dresden.bsky.social covered in smoke from earby wildfires. So this really is the new normal, huh...😒 #climatechange
Finally some feature processing: I personally find getting a feeling for some #unsupervised #ML parameters hard, e.g., in dimensionality reduction/ #clustering. Click yourself through the parameters and see what happens interactively! Needless to say, the algorithms are extensible⬆️📈
Shapes 📐: The new clusters-plotter automatically assigns a fitting colormap to "categorical" features, which is neat when visualizing, e.g. skan (skeleton-analysis.org) #skeletons: Skeleton ids are then automatically assigned fitting colors - no need for manual selections!
With a little help from @stefanhahmann.bsky.social , tracking data is now supported, too. Goes to show the frame-highlighting for #timelapse ⌛data in the new version:
More supported layers: A major upgrade allows to interact with all sorts of @napari.org layer types: Labels🏷️, Points✴️, Vectors🏹, Shapes🔻, Tracks🔀 and even Surfaces⚫! Particularly cool about Surfaces: @mazoc.bsky.social's histogram view leaves unselected areas transparent! 👉 pypi.org/project/biap...
FINALLY just released the @napari.org clusters-plotter 0.9.0 🤗 pypi.org/project/napa... The latest version comes with a complete overhaul of the entire codebase to make #microscopy data introspection and browsing even more accessible, intuitive and versatile! #bioimaging Let's have a look 👇🧪💻🔬
