We’re looking for a way to version and catalogue self-trained deep learning models (training data, code revisions, configs, etc.) from our Tissue Concepts/Medical Multitask Modelling family of foundation models.

We briefly …

github.com/FraunhoferME...

#datascience, #computationalpathology
1/3

🏥 Wie kann #KI das Klinikpersonal während komplexer Eingriffe unterstützen? Im Projekt SmartHospital.NRW entsteht ein innovativer, sprach- und gestengesteuerter Assistent für die Radiologie – entwickelt von der UME Essen, @fraunhoferiais.bsky.social, #FraunhoferMEVIS und weiteren Partnern.

How to unlock the digital sculpture inside the metal print “IntraBeing: Expanded Feet” up at Ctrl Gallery through 2/8/25

#mriart #artscience #ar #nerves #magnetic #flesh #atoms #innerlandscapes #fraunhofermevis #steamimaging #joteva #postdigital #feetnerves

Big shoutout to @tillnitus.bsky.social (who did the lion’s share of the work 🦁), to our co-authors Raphael Schäfer, Henning Höfener, Friedrich Feuerhake, Dorit Merhof, and Fabian Kießling and to the whole team at FraunhoferMEVIS! Special thanks to the reviewers for their constructive feedback!
🧵5/5

You can already find the base model on GitHub (github.com/FraunhoferME...), the pathology-specific model will follow shortly.

🧵4/5

- Extracts transferable features validated across multiple clinical centers 🏥🏥

Model&Code: github.com/FraunhoferME...

Full paper: doi.org/10.1038/s435...

(this is kind of a repost which did not make it to bluesky yet)