Today, I will be attending the #STATBreakthrough West in San Francisco at the Commonwealth Club. Lots of great talks about AI and personal journeys.

Given the concerns about AI writing, be aware that Pangram has a super high false positive rate! It frequently rates my personal writing as 100% LLM written!

Just announced: ai.google/gemini-for-s...

📣 New Paper Alert! Generalist physicians need AI literacy just as much as radiologists do. When I set out to build a new AI course, I couldn't find a medical curriculum guide that addressed this gap, so I decided to build one. Let me walk you through it. 🧵 1/4 #MedSky #MedEd

🙌 Help us understand what #metadata is needed to make #preclinical #imagingdata FAIR! Thanks to your feedback, we will support the development of PRISM - a metadata model for sharing Preclinical Imaging Datasets. @gideproject.bsky.social ⏱️ 20 minutes 🗓️ May 31 www.eurobioimaging.eu/news/collabo...

#HappyMicroscopyMonday #microscopycommunity - Join the BINA Builders for a panel and discussion on the shift from proprietary systems to community-driven innovation. Thursday, May 21! Learn more: buff.ly/pMShl9t

Seth Levine gave a great talk on the art of clustering at ODSC East that really digs in the practical realities of getting the most out of your data. www.youtube.com/watch?v=LtZ3...

Good ML code comes from people who fundamentally believe that ML is a trickster god and wants them to fail

New release of the TrackMate single particle tracking plugin for ImageJ by @jytinevez.bsky.social looks amazing! forum.image.sc/t/trackmate-... Loads of new features, most excitingly to me deep learning segmentation algorithms including Omnipose (for cells), and Spotiflow (for spots) #bioimaging

A holiday demonstration of the different penetration depths of red and blue light into tissue. #optics

Thank you editors for improving the readability of our paper by removing acronyms. Now we grow our cluster of differentiation 4+ Forkhead box 3+ cells in Roswell Park Memorial Institute media with phorbol 12-myristate 13-acetate. TIL, RPMI media is actually Roswell Park Memorial Institute media

We present Olmo 3, our next family of fully open, leading language models. This family of 7B and 32B models represents: 1. The best 32B base model. 2. The best 7B Western thinking & instruct models. 3. The first 32B (or larger) fully open reasoning model.

@ninaluong.bsky.social is really upping her costume game this year. And she can still work in it!

When #halloween meets #FluorescenceFriday 🎃

🎃A fluorescent turmeric jack-o'-lantern!🎃 I filled a flat-bottom flask with turmeric extract. Jordan Tanen drew the face.🖊️ #FluorescenceFriday #SciArt #SciComm #RealTimeChem #ChemArt

The #OME-Zarr RFC-5 proposal allows you to transform multiple images into the same coordinateSystem. Trying this out with napari at github.com/will-moore/n.... Cool that you can offset T-dimension as well as XYZ!

🎃Welcome to my fluorescent pumpkin patch!🎃 Jack-O-Lanterns are pipette tips filled with 😈 lucifer yellow, fluorescein, or rhodamine 101.🎃 🌝 #FluorescenceFriday #PipetteArt #SciArt #ChemArt #SciComm #HappyHalloween #Halloween 🎃

New #NeurIPS2025 paper: how should we evaluate machine learning models without a large, labeled dataset? We introduce Semi-Supervised Model Evaluation (SSME), which uses labeled and unlabeled data to estimate performance! We find SSME is far more accurate than standard methods.

Going through some Erdős problems that Mark Sellke and Mehtaab Sawhney found had already been solved using GPT5 pro for search. More precisely, these were listed as open, but solutions to them already existed in the literature, but not recognized as such. Quite remarkable. One example: [+]

For this #FluorescenceFriday 🔬 a cryosection of a dissected zebrafish gill, stained with phalloidin (orange) and DAPI (blue-ish). #Zebrafish #DevBio #Microscopy

what do you mean i have to submit an abstract now for a presentation in april?

Today for #FluorescenceFriday I’m sharing a 👻Halloween-themed🎃 neural crest explant “web of cells” from Julia Godinez, a 4th year @ucdavis.bsky.social graduate student in the lab. She is studying mechanisms driving conserved and divergent cranial neural crest migration and differentiation. #DevBio

As part of our monthly meeting, we do a run-down of all* the new #bioimageanalysis tools that have been published in the last couple of weeks on BioRxiv**, so I thought I'd start sharing here as well. *I will miss some, apologies if I miss your paper

Stages of CV research work 1. Denial 2. Anger 3. Bargaining 4. Depression 5. Manual data labeling 6. ??? 7. PROFIT

Big news! Our week-long #bioimageanalysis bootcamp is BACK! Learn how to think about bioimage analysis like the pros do, learn what corrections you should (and shouldn't!) apply to your images, learn open source tools, and more. Apply NOW at broad.io/BAB3 for our January 2026 course!

I found these immune cells just finished mitosis but their midbody is still yet to be resolved. How cool? If you zoom in to the midbody, you can see its classical shape just like in the textbooks. I love these little details.

The future of High-Content Imaging has arrived. -> Ultra-fast 3D imaging of multi-well plates at high resolution using an air objective! See previous posts and the new miOPM preprint for details: doi.org/10.1101/2025...

Threw away the pellet

#SpatialOmics MCMICRO Multiple Choice Microscopy pipeline mcmicro.org Multi-channel whole-slide images▶️ Single-cell data + Spatial Feature Tables Multimodal (CyCIF CODEX mIHC) registration + HE IMC MIBI #NatMethods 2022 www.nature.com/articles/s41...

The connectome of the entire male Drosophila central nervous system is out!! Acquired using FIB-SEM @ 8x8x8nm, we aligned 7.5 mio 2D images aligned into one seamless 3D volume, which was used to segment & proofread neurons and synapses. @hhmijanelia.bsky.social www.biorxiv.org/content/10.1...

Happy #FluorescenceFriday! Our #CellLineOfTheWeek is Nucleoporin Nup153! 🧫 Our stem cell lines and plasmids, illuminating specific cell structures, are available for just the cost of shipping to support scientific discovery. 📦 Distributed by @coriellinstitute.bsky.social & @addgene.bsky.social

For #FluorescenceFriday, α-spectrin (red) and lipid droplets (green) are shown in adipocytes 🔬 #Microscopy #CellBiology

After many years of hm-ing and haw-ing about why some of the fusion proteins I’ve cloned showed cytoplasmic background I’ve finally found the culprit

Happy #FluorescenceFriday! This is a maximum intensity projection of depth shaded actin (🌈) in a section from an E9.5 🐭 neural tube (NT). 🔬 by postdoc @christinaadaly.bsky.social 👩‍🔬 🧪 Image shows the floor plate and lumen of the developing NT. #SciArt #DevBio #DevNeuro

🎃💀Boys & girls of every age. Wouldn't you like to see something strange?💀🎃 🖌️🎨I painted Jack Skellington with E. coli expressing the small Ultra-Red Fluorescent Protein (smURFP).🧬🦠🧫 #TheNightmareBeforeChristmas #JackSkellington #SciArt #AgarArt #SciComm

Keynote at #COLM2025: Nicholas Carlini from Anthropic "Are language models worth it?" Explains that the prior decade of his work on adversarial images, while it taught us a lot, isn't very applied; it's unlikely anyone is actually altering images of cats in scary ways.

Now that I can say it publicly - huge congratulations to @michelleitano.bsky.social , winner of this year's @bioimagingna.bsky.social Michael W Davidson award for "Community contributions, in any of the areas of Engagement, Education, Mentorship, and Standardization". So well deserved! #GBIxBINA2025

📄🔈✨ Deep learning is an empirical science, but we rely on basic empirical methods. What might a better foundation—a simple theory—for empirical work look like? @kyunghyuncho.bsky.social, He He, and I move towards one in "Hyperparameter Loss Surfaces Are Simple Near their Optima" at #COLM2025! 🧵1/9

🚨Updated: "How far can we go with ImageNet for Text-to-Image generation?" TL;DR: train a text2image model from scratch on ImageNet only and beat SDXL. Paper, code, data available! Reproducible science FTW! 🧵👇 📜 arxiv.org/abs/2502.21318 💻 github.com/lucasdegeorg... 💽 huggingface.co/arijitghosh/...

💥 DeepInverse v0.3.5 is out! 💥 docs: deepinv.github.io/deepinv/ github: github.com/deepinv/deep...

Segment large images without tiling artifacts: sharing our work that should have been presented at ICCV in 2 weeks - the brilliant first author Elena can’t go because of visa issues. The paper: arxiv.org/abs/2503.19545 1/🧵

Incredible video. I suspect I will end up watching every video on the channel.

Playing around with how to represent volume in #Snouty light sheet data. Here is a dividing hiPSC with endogenously tagged histone in all three orthogonal projections. Bluer is further away. If anything, it distracts from politics.

Check out our new RAEFISH spatial transcriptomics technology published in Cell! This work represents the first time that transcripts from more than 20,000 genes were directly imaged in situ with any technology, and the first time numerous gRNAs were directly probed in an image-based CRISPR screen.

Our Scientists/Artist Alex Ritter (Insta @cells_invitro) took us on his journey from imaging immune cells using high end microscopy to modelling using glass blowing technique, including some new "glow in the dark" pieces using phosphorescent material 🤯 www.toledocellulart.org #SciArt #Microscopy

A good day to remember John Gurdon’s school report from his biology master at Eton

My former team at JAX sent two people to SciPy this year and they teamed up to teach a workshop on scaling-up deep learning inference to large-scale bioimage data - it's finally on youtube! if you need an intro to the dask/zarr world, this is a great starting point. www.youtube.com/watch?v=QCWu...

What an incredible trackathon! After two weeks of intense discussions and happy coding (and a lot of groundwork beforehand by @msschwartz21.bsky.social and Caroline Malin-Mayor), this lovely community emerged with a Graph Exchange File Format that is ready to use! Congrats, everyone! 🎉

Job alert! We are looking for a bioimage analysis to work in an image analysis core facility in beautiful Paris. Can I ask you to share this opportunity with your networks? See also on the forum: forum.image.sc/t/research-e...

Fwd: Fwd: “I hope this email finds you well”