carsten marr (@marrcarsten) 's Twitter Profile
carsten marr

@marrcarsten

Analyzing and modeling single cells in biomedical data

ID: 1011603763428233217

linkhttps://www.helmholtz-munich.de/aih/index.html calendar_today26-06-2018 13:35:06

109 Tweet

644 Followers

130 Following

Bastian Grossenbacher-Rieck (@pseudomanifold) 's Twitter Profile Photo

📢 Our (Salome Kazeminia, carsten marr) paper on topologically regularized multiple instance learning (MIL) for rare anemia disorder classification is on arXiv! arxiv.org/abs/2307.14025 1/8

Sophia J. Wagner (@sophiajwagner) 's Twitter Profile Photo

⚡️Want to save compute when training your feature extractor for histopathology?🔬💻#CPath Study on 1(!) GPU for 2 CRC tasks led by our fantastic master student Benedikt Roth, co-supervised with Valentin Koch pre-print: arxiv.org/abs/2401.04720 code: github.com/beneroth13/din… 1/

⚡️Want to save compute when training your feature extractor for histopathology?🔬💻#CPath
Study on 1(!) GPU for 2 CRC tasks led by our fantastic master student <a href="/roth_benedikt13/">Benedikt Roth</a>, co-supervised with <a href="/ValentinKoch_/">Valentin Koch</a> 
pre-print: arxiv.org/abs/2401.04720
code: github.com/beneroth13/din…
1/
Sophia J. Wagner (@sophiajwagner) 's Twitter Profile Photo

Takeaways: - finetuning can be achieved with low resources (we used 1 GPU only + batch size 32 was sufficient for training ViT-g) - even for small datasets finetuning strong CV foundation models might be better than histo-pre-trained feature extractors 3/

Takeaways: 
- finetuning can be achieved with low resources (we used 1 GPU only + batch size 32 was sufficient for training ViT-g)
- even for small datasets finetuning strong CV foundation models might be better than histo-pre-trained feature extractors
3/
Ewa Szczurek (@ewa_szczurek) 's Twitter Profile Photo

My new lab at the Institute of AI for Health at Helmholtz Munich | @HelmholtzMunich is looking for postdocs! Great projects in AI for discovery of antimicrobial peptides and Computational oncology. RT or apply! jobs.helmholtz-muenchen.de/jobposting/2b4…

Bjoern Eskofier (@bjoerneskofier) 's Twitter Profile Photo

Interested in #Foundation #Models in #Health / #Medicine? -> Consider stopping by: April 10, Helmholtz Munich | @HelmholtzMunich 3 #keynotes, an exciting scientific program, and networking opportunities await. Registration / abstract submission: events.hifis.net/event/1221/ Fabian Theis carsten marr

Interested in #Foundation #Models in #Health / #Medicine?

-&gt; Consider stopping by: April 10, <a href="/HelmholtzMunich/">Helmholtz Munich | @HelmholtzMunich</a>

3 #keynotes, an exciting scientific program, and networking opportunities await. Registration / abstract submission: events.hifis.net/event/1221/

<a href="/fabian_theis/">Fabian Theis</a> <a href="/MarrCarsten/">carsten marr</a>
Bastian Grossenbacher-Rieck (@pseudomanifold) 's Twitter Profile Photo

🎉ICLR 2026: Simplicial Representation Learning w/ Neural k-Forms #Geometry #MachineLearning #ICLR2024 📜openreview.net/forum?id=Djw0X… TL;DR: Learnable differential forms = ꜰᴇᴀᴛᴜʀᴇ ᴍᴀᴘꜱ for combinatorial complexes (like graphs), enabling classification + more! 🧵1/n

🎉<a href="/iclr_conf/">ICLR 2026</a>: Simplicial Representation Learning w/ Neural k-Forms

#Geometry #MachineLearning #ICLR2024

📜openreview.net/forum?id=Djw0X…

TL;DR: Learnable differential forms = ꜰᴇᴀᴛᴜʀᴇ ᴍᴀᴘꜱ for combinatorial complexes (like graphs), enabling classification + more!

🧵1/n
Bastian Grossenbacher-Rieck (@pseudomanifold) 's Twitter Profile Photo

🎉 Stoked that Ernst Röell's first Ph.D. work "Differentiable Euler Characteristic Transforms for Shape Classification" was accepted ICLR 2026 #ICLR2024 👉A new, super fast topological layer (based on Euler Characteristic Transform) #geometry #topology #MachineLearning 🧵1/n

Manuel Tran (@tranmanuel) 's Twitter Profile Photo

I am thrilled to announce HistoGPT, a vision language foundation model that generates highly accurate pathology reports from gigapixel whole slide images. medRxiv: medrxiv.org/content/10.110… GitHub: github.com/marrlab/HistoG… Hugging Face: huggingface.co/marr-peng-lab/…

I am thrilled to announce HistoGPT, a vision language foundation model that generates highly accurate pathology reports from gigapixel whole slide images.

medRxiv: medrxiv.org/content/10.110…
GitHub: github.com/marrlab/HistoG…
Hugging Face: huggingface.co/marr-peng-lab/…
Eugenio F. Fornasiero (@euforna) 's Twitter Profile Photo

Proud of the first chapter, hopefully of many, of our collaboration with Jungmann Lab published in Cell where SUM-PAINT was developed for imaging neurons. Kudos to Eduard Unterauer Kristina Mahipal Ganji Luciano A. Masullo carsten marr Felipe Opazo. sciencedirect.com/science/articl…

Proud of the first chapter, hopefully of many, of our collaboration with <a href="/JungmannLab/">Jungmann Lab</a> published in <a href="/CellCellPress/">Cell</a> where SUM-PAINT was developed for imaging neurons. Kudos to Eduard Unterauer <a href="/jekriste/">Kristina</a> <a href="/mahipal_ganji/">Mahipal Ganji</a> <a href="/l_masu/">Luciano A. Masullo</a> <a href="/MarrCarsten/">carsten marr</a> <a href="/FelipeOpaz_o/">Felipe Opazo</a>. sciencedirect.com/science/articl…
Helmholtz Munich | @HelmholtzMunich (@helmholtzmunich) 's Twitter Profile Photo

#PositionPaper 🇩🇪 Experts offers solutions to navigate the complexities of using #patient #data in #medical #research, addressing privacy concerns while maximizing #AI's potential. 👉Read more about their recommendations t1p.de/yw7a6 #HelmholtzMunich carsten marr

#PositionPaper 🇩🇪
 
Experts offers solutions to navigate the complexities of using #patient #data in #medical #research, addressing privacy concerns while maximizing #AI's potential.

👉Read more about their recommendations t1p.de/yw7a6

#HelmholtzMunich <a href="/MarrCarsten/">carsten marr</a>
Valentin Koch (@valentinkoch_) 's Twitter Profile Photo

Our innovation grant-winning study on AI-based OCT analysis post-PCI is on medRxiv! DeepNeo is a tool for neointima & stent analysis with a GUI (⬇️) ! It classifies neoint. w same acc as experts + strong segmentation of lumen, stent and neoint.🔬 1/6 medrxiv.org/cgi/content/sh…

Our innovation grant-winning study on AI-based OCT analysis post-PCI is on medRxiv! DeepNeo is a tool for neointima &amp; stent analysis with a GUI (⬇️) ! It classifies  neoint. w same acc as experts + strong segmentation of lumen, stent and neoint.🔬 1/6
medrxiv.org/cgi/content/sh…
Lennart Kuck, PhD (@kucklennart) 's Twitter Profile Photo

Difference in Piezo1-dependent Ca2+-influx, so what? To show "what", we teamed up with carsten marr's lab, adopting their custom-developed RBC morphology analysis algorithm. Ca2+-influx causes K+-efflux and loss of water from RBC - cells shrink. 8/n

Difference in Piezo1-dependent Ca2+-influx, so what? To show "what", we teamed up with <a href="/MarrCarsten/">carsten marr</a>'s lab, adopting their custom-developed RBC morphology analysis algorithm. Ca2+-influx causes K+-efflux and loss of water from RBC - cells shrink.
8/n
Helmholtz Munich | @HelmholtzMunich (@helmholtzmunich) 's Twitter Profile Photo

Introducing HistoGPT: A Foundation Model for AI-Powered Pathology Reports. 💡A new AI tool with the potential to transform dermatopathology by automating process of generating high-quality pathology reports from histology images. 🔗t1p.de/83o0a

Introducing HistoGPT: A Foundation Model for AI-Powered Pathology Reports. 

💡A new AI tool with the potential to transform dermatopathology by automating process of generating high-quality pathology reports from histology images. 

🔗t1p.de/83o0a