🔥 So much in this paper: CT RATE (one of the largest public chest CT/text report datasets), CT-CLIP (a 3D chest CT foundation model) and CT-CHAT, a conversational model building on CT-CLIP @ibrahimethem.bsky.social

🚨 Agentic Systems in Radiology Everyone’s talking about agents🕵️‍♀️🕵️‍♂️ — but what happens when you bring them into real clinical workflows? Our new preprint explores the design, applications, challenges, and evaluation of LLM-based agents and agentic workflows in radiology 👇 arxiv.org/abs/2510.09404

🚨 Agentic Systems in Radiology Everyone’s talking about agents🕵️‍♀️🕵️‍♂️ — but what happens when you bring them into real clinical workflows? Our new preprint explores the design, applications, challenges, and evaluation of LLM-based agents and agentic workflows in radiology 👇 arxiv.org/abs/2510.09404

💥 We unveil our paper accepted at the #ACL2025 Main Conference: Automated Structured Report Generation Let's revisit automated radiology report generation for CXR. Free-form reports make it hard for AI systems to learn accurate generation, and even harder to evaluate. 🧵👇 @StanfordAIMI @hopprai

👍 If you're interested in #LLMs in #radiology, this is a recommended read! 💯 While the article focuses primarily on LLMs, as the authors recommended, "Keep an eye on [large multimodal models]". 👉 pubs.rsna.org/doi/10.1148/... #RadiologyAI #AIStrategy #LMMs

When reading AI benchmarks, aside from the fact that many of the AIs are (accidentally or on purpose) trained on the test set, many tests are just bad. MMLU likely maxes out at 90% or so because so many of the questions in it are just wrong. It is also uncalibrated in difficulty across questions.

LLMs formally pass the Turing test arxiv.org/abs/2503.23674

lot of folks talking about MCP and developers rushing to support it - but it does seem a bit like hype-y. interesting perspective: hardcoresoftware.learningbyshipping.com/p/230-mcp-it...

that's a first :O

A quarter of startups in YC’s current cohort have codebases that are almost entirely AI-generated

Chaired two insightful sessions at #ECR2025 today! "Standardization and Reporting in AI Research" with experts Hans Reitsma Mike Klontzas Annika Reinke "How to Use ChatGPT for Academic and Administrative Tasks" Andreas S. Brendlin @cxbln.bsky.social and Ghizlane Lembarki @myesr.bsky.social

still thinking about this interaction with Claude (Oct '24) from time to time

r1ing away on an ordinary machine .. straight out of Kahnemans lesser known book "Thinking .. mostly slow"

For Germans unable to vote locally, it is easy to partake, here's how: www.bundesregierung.de/breg-de/schw...

The Illustrated DeepSeek-R1 Spent the weekend reading the paper and sorting through the intuitions. Here's a visual guide and the main intuitions to understand the model and the process that created it. newsletter.languagemodels.co/p/the-illust...

another excellent blog post by @chiphuyen.bsky.social, this time on #agents huyenchip.com/2025/01/07/a...

People: please don't ML for the sake of ML. I keep seeing manuscripts using fancy machine learning on brain-imaging data, where, in my opinion (having processed a lot of brain-imaging data), the method is way too complex for the richness of the data. Fancier is not better per se

The promise and perils of synthetic data

"productization requires [standardization in development and deployment], which is antithetical to research. [...] PhDs are supposed to come up with innovative ideas, validate these ideas, report the findings to the community by writing papers and then move on." [slightly edited to fit post limits]

A market research team's dream: gathering real-world use cases and improvement opportunities directly from natural product usage from: www.anthropic.com/research/clio

NEW PREPRINT A detailed overview of 32 popular predictive performance metrics for prediction models arxiv.org/abs/2412.10288

VLMs go MoE ✨ DeepSeek AI dropped three new commercially permissive vision LMs based on SigLIP encoder and their DeepSeek-MoE decoder 🐳 the models come in 1.0B, 2.8B and 4.5B active params 🥹 models seem to catch up with state-of-the-art with less active parameters! huggingface.co/collections/...

the only way to influence this future is to shape it — h/t @adamrodmanmd.bsky.social @fchollet.bsky.social

Join us tomorrow for the session "Future Frontiers in Health AI: Emerging Technologies and Innovations" at #AIPlusHealth24 #RadSky #Radiology #StanfordAIMI

We thank our amazing reviewers for their time and expertise. This work cloud highlights the reviewers who contribute to @radiology-ai. Thank you, peer reviewers! #RSNA2024 #radiology #AI

Tree in bud nodularity on a CXR (via barium aspiration) 🫁 #radiology #radsky #chestrad @laurengroner.bsky.social @chestradiologist.bsky.social @mmestas.bsky.social @avrahamcoopermd.bsky.social

I do dabble in Multi-1Ical Imaging occasionally

who needs to keep track of all the profiles of conferences in #AI and #ML on 🦋? go.bsky.app/SZEr8Jf

And chatbots aren't moral agents nor have ownership of the text they output, so they don't need to be credited. The proper thing to do when using a chatbot is to disclose it, for transparency, rather than cite it for attributing credit or signaling credibility. 4/

my (and probably many others’) entry point to ML research was this man's course, using Octave back then in 2016. so extremely honored to share an author list now with Prof. @andrewyng.bsky.social ! thank you

Highlights from the President's Address: Building Intelligent Connections from @curtlanglotz.bsky.social : Anyone who works with AI knows -- machine intelligence is different---not better---than human intelligence. #RSNA24 #RadSky

My deep learning course at the University of Geneva is available on-line. 1000+ slides, ~20h of screen-casts. Full of examples in PyTorch. fleuret.org/dlc/ And my "Little Book of Deep Learning" is available as a phone-formatted pdf (nearing 700k downloads!) fleuret.org/lbdl/

In times of #RSNA24 this #radiology feed should have more than currently 40 followers: bsky.app/profile/did:...

We are excited to share the launch of our company - Cognita! We are working to build the future of radiology through multi-modal AI systems with a great group of founders @loublanks.bsky.social, @akshay-chaudhari.bsky.social, and I, and advisors Ajit Singh, Chris Re, and @curtlanglotz.bsky.social.

We are excited to share the launch of our company - Cognita! We are working to build the future of radiology through multi-modal AI systems with a great group of founders @zhihongc.bsky.social, @akshay-chaudhari.bsky.social and I, and advisors Ajit Singh, Chris Re, and @curtlanglotz.bsky.social 1/2

Underresearched subject in this recent JAMA study "AI use [in radiology practice] was significantly associated with increased odds of burnout (OR 1.20; 95% CI, 1.10-1.30), primarily driven by its association with [emotional exhaustion] (OR, 1.21; 95% CI, 1.10-1.34)." (corr≠caus) #Radsky #AI

Medically adapted foundation models (think Med-*) turn out to be more hot air than hot stuff. Correcting for fatal flaws in evaluation, the current crop are no better on balance than generic foundation models, even on the very tasks for which benefits are claimed. arxiv.org/abs/2411.04118

"Explainability" and "interpretability" are often being used inconsistently, so this is a good reference

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...

Tip for doctors joining #Medsky: Add a #Medsky label to your account: bsky.app/profile/did:... Then you will show up on relevant feeds, which you can pin to your home, such as Radiology #Radsky: bsky.app/profile/meds... Engagement-ranked high-yield general Medsky: bsky.app/profile/meds...

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

yet

Imaging guided interventions imply the existence of imaging misguided interventions

Super excited to introduce Halo: A beginner's guide to DIY health tracking with wearables! 🤗✨ Using an $11 smart ring, I'll show you how to build your own private health monitoring app. From basic metrics to advanced features like: - Activity tracking - HR monitoring - Sleep analysis and more!

There is a lot of energy going into fine-tuning models, but specialized medical AI models lost to their general versions 38% of the time, only won 12%. Before spending millions on specialized training, might be worth exploring what base models can do with well-designed prompts.

Based on seeing lots of companies, 98% of what people are calling AI agents are not what the AI labs would call agents. They are usually structured document retrieval systems with a prompt or two for summaries, there is very little control or decision-making given to the AI, very little AI planning

Hello World!