Keno Bressem

Image from article in Radiology: Artificial Intelligence

MRSegmentator accurately segmented 40 anatomical structures on #MRI and #CT across 3 external datasets https://doi.org/10.1148/ryai.240477 @kbressem.bsky.social #segmentation #MachineLearning #AI

Image from article in Radiology: Artificial Intelligence

New comprehensive whole-body MRI #segmentation tool: MRSegmentator https://doi.org/10.1148/ryai.240477 @kbressem.bsky.social #MRI #MachineLearning #AI

Image from article in Radiology: Artificial Intelligence

MRSegmentator accurately segments 40 anatomic structures on MRI and generalized to CT https://doi.org/10.1148/ryai.240477 @kbressem.bsky.social #CT #radiology #AI

Image from article in Radiology: Artificial Intelligence

MRSegmentator accurately segmented 40 anatomical structures on #MRI and #CT across 3 external datasets https://doi.org/10.1148/ryai.240477 @kbressem.bsky.social #radiology #MachineLearning #AI

Cybersecurity Threats and Mitigation Strategies for Large Language Models in Health Care | Radiology: Artificial Intelligence “Just Accepted” papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and p...

📢 Check out our new article on cybersecurity threats posed by LLMs @radiology-ai.bsky.social

🔗 doi.org/10.1148/ryai...

@judywawira.bsky.social @alitejanimd.bsky.social @shahriarfaghani.bsky.social @bardiakh.bsky.social @cxbln.bsky.social @kbressem.bsky.social @felixbusch.bsky.social @Lisa @Mana

Very impressive. aistudio.google.com/live truly feels like SciFi.

First open-weights for OpenAI-o1-like reasoning model! QwQ from the Qwen team is a 32B model that beats OpenAI O1 mini and competes w/ O1 preview and is available under Apache 2.0 on Hugging Face! 🤯

Regular reminder to science Bluesky: do not submit your work to MDPI or Frontiers journals

This is very similar to what we also found in a recent study. As long as we use openly available medical data and don't incorporate large amounts of currently inaccessible medical data, the generalist models will always be the better choice.

A statistical approach to model evaluations A research paper from Anthropic on how to apply statistics to improve language model evaluations

As people move from proof-of-principle to more serious evaluations of LLMs, including assessment of statistical significance, insights from decades of statistical research can be useful.

Nice blog post & paper with practical recommendations for #LLM evaluation.

www.anthropic.com/research/sta...

A small starter pack for accounts in the radiology/AI space (or prolific in either of those fields) - happy to add more! go.bsky.app/9Drtasz #RadSky #Radiology #AI