Light-sheet 3D imaging paired with ML is accelerating whole-brain and organ-level reconstructions. Tools like csbdeep and StarDist, together with HPC-backed pipelines, let bioimaging teams move from manual ROI review to reproducible, batch-ready cell and organelle segmentation. Great teaser.
November 8, 2026 at 2:51 AM
Light-sheet 3D imaging paired with ML is accelerating whole-brain and organ-level reconstructions. Tools like csbdeep and StarDist, together with HPC-backed pipelines, let bioimaging teams move from manual ROI review to reproducible, batch-ready cell and organelle segmentation. Great teaser.
🧙 With a wizard that inspects your images & labels, previews them and flags issues before training, selects an AI model (nnUNet style), analyses object shapes to recommend the best representation (Cellpose flows, #StarDist, contours, distance maps, affinities…), and suggests a training budget.
November 11, 2026 at 3:32 PM
🧙 With a wizard that inspects your images & labels, previews them and flags issues before training, selects an AI model (nnUNet style), analyses object shapes to recommend the best representation (Cellpose flows, #StarDist, contours, distance maps, affinities…), and suggests a training budget.
Excited to share my new study where we measured temporal evolution of vesicle morphometrics, such as area, contact curvature, local membrane curvature during budding. The data was analyzed by exploiting StarDist as a model to find daughter vesicles. (1/2)
Evolution of Vesicle Shape Parameters Based on A Deep-Learning Based Image Analysis Framework Over Extended Time Periods https://www.biorxiv.org/content/10.64898/2026.11.03.764022v1
November 5, 2026 at 3:11 PM
Excited to share my new study where we measured temporal evolution of vesicle morphometrics, such as area, contact curvature, local membrane curvature during budding. The data was analyzed by exploiting StarDist as a model to find daughter vesicles. (1/2)
Led by António Brito and Bruno Saraiva, with Dominik Alwardt, Beatriz de P. Mariz, Sérgio Filipe and @pinholab.bsky.social, at @itqbnova.bsky.social. Built on @napari.org: StarDist, Cellpose and U-Net, plus a retrainable classifier =)
September 15, 2026 at 9:18 AM
Led by António Brito and Bruno Saraiva, with Dominik Alwardt, Beatriz de P. Mariz, Sérgio Filipe and @pinholab.bsky.social, at @itqbnova.bsky.social. Built on @napari.org: StarDist, Cellpose and U-Net, plus a retrainable classifier =)
🔬🚀 Out now in Communications AI & Computing: mAIcrobe, our open-source napari plugin for high-throughput bacterial image analysis. Segmentation, morphological profiling and cell cycle classification, all in one place 👇
doi.org/10.1038/s44488-026-00022-y
doi.org/10.1038/s44488-026-00022-y
September 15, 2026 at 9:18 AM
🔬🚀 Out now in Communications AI & Computing: mAIcrobe, our open-source napari plugin for high-throughput bacterial image analysis. Segmentation, morphological profiling and cell cycle classification, all in one place 👇
doi.org/10.1038/s44488-026-00022-y
doi.org/10.1038/s44488-026-00022-y
Pipeline release! nf-core/sopa v1.0.1 - 1.0.1!
Please see the changelog: https://github.com/nf-core/sopa/releases/tag/1.0.1
Please see the changelog: https://github.com/nf-core/sopa/releases/tag/1.0.1
Release 1.0.1 · nf-core/sopa
Fixes
Set writable env vars for StarDist and Proseg segmentation (#35)
Ensure the tests can't silently fail (#29)
Updates
Support the image_scale parameter (#30)
Update the spaceranger module (#...
github.com
August 25, 2026 at 10:08 AM
Pipeline release! nf-core/sopa v1.0.1 - 1.0.1!
Please see the changelog: https://github.com/nf-core/sopa/releases/tag/1.0.1
Please see the changelog: https://github.com/nf-core/sopa/releases/tag/1.0.1
Thierry Neuville "Usun, et ka temal tuleb veel raskeid hetki – ja just siis peame meie olema valmis need võimalused maksimaalselt ära kasutama.“
ralli.ee/thierry-neuv...
ralli.ee/thierry-neuv...
Thierry Neuville ei lase Evansi heast stardist end heidutada: „Küll temalgi tuleb veel raskeid hetki“ - Ralli.ee
Thierry Neuville tunnustab Elfyn Evansi tugevat hooaja algust, kuid usub, et raskemad hetked ootavad ees – ja siis tuleb Hyundai'l võimalus.
ralli.ee
March 30, 2025 at 1:50 PM
Thierry Neuville "Usun, et ka temal tuleb veel raskeid hetki – ja just siis peame meie olema valmis need võimalused maksimaalselt ära kasutama.“
ralli.ee/thierry-neuv...
ralli.ee/thierry-neuv...
Munster alustas päeva karjääri esimese katsevõiduga – Neuville Hyundail tehnilised probleemid
ralli.ee/ss10-munster...
ralli.ee/ss10-munster...
SS10: Gregoire Munster alustas päeva karjääri esimese WRC katsevõiduga - Neuville Hyundail tehnilised probleemid - Ralli.ee
Laupäevase päeva esimesel katsel tekkisid valitseva meistri, Thierry Neuville masinal tehnilised probleemid. Belglase masin läks kohe stardist ...
ralli.ee
January 25, 2025 at 8:47 AM
Munster alustas päeva karjääri esimese katsevõiduga – Neuville Hyundail tehnilised probleemid
ralli.ee/ss10-munster...
ralli.ee/ss10-munster...
Konstaterar att det börjar bli läge för ett tilltryck av Tobias Hadins uppskattade zombierollspel "Skjut dom i huvudet". Låga lager hos våra distributörer Stardist och Omnix samt även i Vildhallons webshop. Kul som bara den!
March 19, 2025 at 7:30 PM
Konstaterar att det börjar bli läge för ett tilltryck av Tobias Hadins uppskattade zombierollspel "Skjut dom i huvudet". Låga lager hos våra distributörer Stardist och Omnix samt även i Vildhallons webshop. Kul som bara den!
While it has its limitations, I believe it offers a strong framework for TIL analysis—both microscopically by pathologists and computationally—with excellent interpretability. As a bonus, we incorporated a StarDist model trained specifically for IHC slides, designed to run seamlessly in QuPath.
November 25, 2024 at 7:16 AM
While it has its limitations, I believe it offers a strong framework for TIL analysis—both microscopically by pathologists and computationally—with excellent interpretability. As a bonus, we incorporated a StarDist model trained specifically for IHC slides, designed to run seamlessly in QuPath.
Next, we applied a pre-trained StarDist deep learning model in QuPath for cell detection and used the extracted cell features to train a binary classifier distinguishing TILs from other cells. To evaluate our TILs assessment pipeline, we calculated [3/5 of https://arxiv.org/abs/2504.16979v1]
April 25, 2025 at 6:16 AM
Next, we applied a pre-trained StarDist deep learning model in QuPath for cell detection and used the extracted cell features to train a binary classifier distinguishing TILs from other cells. To evaluate our TILs assessment pipeline, we calculated [3/5 of https://arxiv.org/abs/2504.16979v1]
Love to hear more about testing. I'm always skeptical when I hear "new method B" outperforms previous method "A" because it often turns out "A" was not run optimally. Seems reasonable in an unsupervised or general mode this new method could be better than generalist Stardist/Cellpose but ...
July 12, 2025 at 1:49 PM
Love to hear more about testing. I'm always skeptical when I hear "new method B" outperforms previous method "A" because it often turns out "A" was not run optimally. Seems reasonable in an unsupervised or general mode this new method could be better than generalist Stardist/Cellpose but ...
Q2 Did you use the tensorflow version of your semantic segmentation notebooks? Any plans to move to Pytorch? (I know you guys set up a lot of things on the cloud, but for Windows Users the tensor flow (and stardist) installation is becoming increasingly convoluted.
May 1, 2025 at 7:03 PM
Q2 Did you use the tensorflow version of your semantic segmentation notebooks? Any plans to move to Pytorch? (I know you guys set up a lot of things on the cloud, but for Windows Users the tensor flow (and stardist) installation is becoming increasingly convoluted.
I was exploring pixi a bit today, but also wondering if you can handle transitive dependencies with conda and pip-compile from pip-tools?? See thoughts on this thread forum.image.sc/t/difficulty...
Difficulty installing Stardist + TensorFlow + GPUtools with anaconda
Exploring this topic a bit today and pixi seems really powerful but also another tool for developer and (perhaps more importantly) end user to have to worry about. Is there a flaw in doing it with a ...
forum.image.sc
November 17, 2024 at 8:32 PM
I was exploring pixi a bit today, but also wondering if you can handle transitive dependencies with conda and pip-compile from pip-tools?? See thoughts on this thread forum.image.sc/t/difficulty...
For DL stuff it depends on the model size and available data to train. A regular nvidia consumer’s card can handle most of the small sized models today, like cellpose and stardist. But of course it will limit large-scale experiments
January 10, 2025 at 8:23 AM
For DL stuff it depends on the model size and available data to train. A regular nvidia consumer’s card can handle most of the small sized models today, like cellpose and stardist. But of course it will limit large-scale experiments
And then once you have 2 you can collect more copies and use them with the stardist currency to add flair to cards. Thats why most of the trainer cards in this deck have pretty sparkles or designs on them. The collecting never ends. The game does not explain the flair system at all though.
November 11, 2024 at 5:15 AM
And then once you have 2 you can collect more copies and use them with the stardist currency to add flair to cards. Thats why most of the trainer cards in this deck have pretty sparkles or designs on them. The collecting never ends. The game does not explain the flair system at all though.
My #BioimageTooloftheDay is an easy one: Cellpose by @computingnature.neuromatch.social.ap.brid.gy and colleagues
An awesome DL-segmentation network to segment cells from microscopy images
www.nature.com/articles/s41...
When we do not use StarDist, we use Cellpose 😀
An awesome DL-segmentation network to segment cells from microscopy images
www.nature.com/articles/s41...
When we do not use StarDist, we use Cellpose 😀
Cellpose 2.0: how to train your own model - Nature Methods
Cellpose 2.0 improves cell segmentation by offering pretrained models that can be fine-tuned using a human-in-the-loop training pipeline and fewer than 1,000 user-annotated regions of interest.
www.nature.com
November 15, 2024 at 4:40 PM
My #BioimageTooloftheDay is an easy one: Cellpose by @computingnature.neuromatch.social.ap.brid.gy and colleagues
An awesome DL-segmentation network to segment cells from microscopy images
www.nature.com/articles/s41...
When we do not use StarDist, we use Cellpose 😀
An awesome DL-segmentation network to segment cells from microscopy images
www.nature.com/articles/s41...
When we do not use StarDist, we use Cellpose 😀
Today I would like to highlight one of our favorite DL-segmentation network: StarDist by @maweigert.bsky.social and colleagues. In my lab we love and use StarDist all the time
github.com/stardist/sta...
link.springer.com/chapter/10.1...
#BioimageTooloftheDay
github.com/stardist/sta...
link.springer.com/chapter/10.1...
#BioimageTooloftheDay
GitHub - stardist/stardist: StarDist - Object Detection with Star-convex Shapes
StarDist - Object Detection with Star-convex Shapes - stardist/stardist
github.com
November 14, 2024 at 3:34 PM
Today I would like to highlight one of our favorite DL-segmentation network: StarDist by @maweigert.bsky.social and colleagues. In my lab we love and use StarDist all the time
github.com/stardist/sta...
link.springer.com/chapter/10.1...
#BioimageTooloftheDay
github.com/stardist/sta...
link.springer.com/chapter/10.1...
#BioimageTooloftheDay
We built StarDist to detect cells in volumetric microscopy and jointly infer closed curved parametric surfaces for each cell. It's an example of an interpretable machine learning approach. More detail in the blog post and preprint.
www.focolab.org/post/surfdis...
www.focolab.org/post/surfdis...
"SurfDist" pre-print posted
We developed an "interpretable machine learning" system to find and characterize blobs in volumetric images, such as microscopy images of cells, something we need to do all the time in our lab. The no...
www.focolab.org
July 14, 2025 at 6:16 PM
We built StarDist to detect cells in volumetric microscopy and jointly infer closed curved parametric surfaces for each cell. It's an example of an interpretable machine learning approach. More detail in the blog post and preprint.
www.focolab.org/post/surfdis...
www.focolab.org/post/surfdis...
Try it yourself! The plugin, documentation, and example datasets are available at:
multichannel-nuclear-analysis.github.io
Special thanks to the StarDist team for their incredible segmentation algorithm that made this possible.
#ImageAnalysis #FIJI #ImageJ #Microscopy
multichannel-nuclear-analysis.github.io
Special thanks to the StarDist team for their incredible segmentation algorithm that made this possible.
#ImageAnalysis #FIJI #ImageJ #Microscopy
Multichannel Nuclear Analysis plugin
An ImageJ/FIJI plugin for analyzing multi-channel fluorescence images with a focus on nuclear segmentation using StarDist. https://multichannel-nuclear-analysis.github.io/
multichannel-nuclear-analysis.github.io
June 7, 2025 at 2:41 PM
Try it yourself! The plugin, documentation, and example datasets are available at:
multichannel-nuclear-analysis.github.io
Special thanks to the StarDist team for their incredible segmentation algorithm that made this possible.
#ImageAnalysis #FIJI #ImageJ #Microscopy
multichannel-nuclear-analysis.github.io
Special thanks to the StarDist team for their incredible segmentation algorithm that made this possible.
#ImageAnalysis #FIJI #ImageJ #Microscopy
The Multi-Channel Nuclear Analysis plugin combines the powerful nuclear segmentation of StarDist (by
Uwe Schmidt and colleagues) with an intuitive interface that makes quantitative fluorescence analysis accessible to everyone, regardless of coding background.
Uwe Schmidt and colleagues) with an intuitive interface that makes quantitative fluorescence analysis accessible to everyone, regardless of coding background.
June 7, 2025 at 2:41 PM
The Multi-Channel Nuclear Analysis plugin combines the powerful nuclear segmentation of StarDist (by
Uwe Schmidt and colleagues) with an intuitive interface that makes quantitative fluorescence analysis accessible to everyone, regardless of coding background.
Uwe Schmidt and colleagues) with an intuitive interface that makes quantitative fluorescence analysis accessible to everyone, regardless of coding background.
#Microscopy folk: If Stardist has stopped working for you since the recent update, reset the tensor flow update site on FIJI!
January 30, 2025 at 5:25 PM
#Microscopy folk: If Stardist has stopped working for you since the recent update, reset the tensor flow update site on FIJI!
As we show in the paper, this strategy can be seamlessly used with other state-of-the-art instance segmentation networks (#StarDist, #Cellpose, #PlantSeg) and the results are equally of high quality. We opted for our own stable light-weight nets:
link.springer.com
link.springer.com
link.springer.com
May 7, 2026 at 9:40 AM
In need of automatic tools for segmenting nuclei in 2D or 3D images? Check out the wonderful [NEUBIASAcademy@Home] Webinar "Introduction to nuclei segmentation with #StarDist"! Pretty neat work by @martweig, @uschmidt83,@ChiguireKun and @SuperResoluSian😊 https://youtu.be/Amn_eHRGX5M
Introduction to nuclei segmentation with StarDist - [NEUBIASAcademy@Home] Webinar
YouTube · NEUBIAS
youtu.be
May 7, 2026 at 9:03 AM
In need of automatic tools for segmenting nuclei in 2D or 3D images? Check out the wonderful [NEUBIASAcademy@Home] Webinar "Introduction to nuclei segmentation with #StarDist"! Pretty neat work by @martweig, @uschmidt83,@ChiguireKun and @SuperResoluSian😊 https://youtu.be/Amn_eHRGX5M
Day 3️⃣ of our @embo.org course on "Advanced methods in bioimage analysis" at @embl.org with "Practical session 3: Tracking" 🔬🦠🔀.
@jytinevez.bsky.social is already shining at the stage! 🕺🪩.
Important update: #TrackMate v8 is coming! 🤗
@jytinevez.bsky.social is already shining at the stage! 🕺🪩.
Important update: #TrackMate v8 is coming! 🤗
September 16, 2025 at 8:10 AM
Day 3️⃣ of our @embo.org course on "Advanced methods in bioimage analysis" at @embl.org with "Practical session 3: Tracking" 🔬🦠🔀.
@jytinevez.bsky.social is already shining at the stage! 🕺🪩.
Important update: #TrackMate v8 is coming! 🤗
@jytinevez.bsky.social is already shining at the stage! 🕺🪩.
Important update: #TrackMate v8 is coming! 🤗