🚨 Published in @Nature today: BOTH of our @GoogleAI @GoogleDeepMind papers on AMIE, our research conversational medical AI 🤖 🩺 @EricTopol - heartfelt thanks for sharing this with the larger community :)

This work is incredibly close to my heart. So grateful to our team @GoogleDeepMind @GoogleResearch & @Yale collaborators. It's been a hard road and a hard-won milestone, and this is just the beginning. The path from a finding like this to the clinic is very long, requiring

An exciting milestone for AI in science: Our C2S-Scale 27B foundation model, built with @Yale and based on Gemma, generated a novel hypothesis about cancer cellular behavior, which scientists experimentally validated in living cells.  With more preclinical and clinical tests,

Learn about user insights from a novel AI agent — “Wayfinding AI”, an early-stage research prototype — that helps find the way to better health information through proactive conversational guidance, goal understanding, and tailored conversations. More → https://t.co/MrjAzt6s6u

The global health workforce is projected to face a shortage of 11 million workers by 2030. We’re exploring how @Google’s AI models could help address this challenge by serving as helpful tools in medical learning environments. More from @GoogleResearch ⬇️

Check out today’s blog post, where we highlight two studies that explore how Google’s AI models can be used as helpful tools in medical learning environments, reimagining the future of education for health professionals. Learn more: https://t.co/qhCf9VvE05

🚨New research from @GoogleResearch @GoogleDeepMind: Towards physician-centered oversight of conversational diagnostic AI. Conversational AI holds promise for improving diagnostic workflows, but safe use must ensure that licensed physicians stay accountable for clinical decisions

Today, Google Research & @GoogleDeepMind introduce g-AMIE, an extension of our diagnostic AI system based on #Gemini 2.0 Flash. It uses a guardrail that prohibits medical advice sharing & instead provides a summary for a physician to review. Learn more: https://t.co/zXc2yWRWLn

Joint with with a great team from @GoogleResearch @GoogleDeepMind incl. @elahevedadi @dgtbarrett @imnatalie_grace Ellery Wulczyn @zhazhir @Romes13 @AdamRodmanMD @hardyshakerman @vivnat @alan_karthi @RyutaroTanno @weballergy @_khaledsaab @JanFreyberg @timstro @dylanslack20 …

LLMs are becoming more capable in health, and we need to think of safe ways to make use of them. In our recent AMIE work, we envision using LLMs to directly scale the physician. We imagine a setting where a physician deploys tens-hundreds of LLMs to gather patient information

Grateful to an incredible team incl: @KhaledSaab11 @JanFreyberg @_cjpark @alan_karthi @RyutaroTanno @apalepu13 @taotu831 @valentinlievin @ckbjimmy @davidstutz92 @vivnat @AdamRodmanMD @timstro @dgtbarrett @SaraM66905 @pushmeet @joelle_barral @dr2w @AvinatanH48021 @ymatias +others!

We detail important limitations in the paper and the results should be interpreted with appropriate caution. Further work is required to safely validate the potential capabilities in prospective settings which we are excited to share more about soon. (6/n)

AMIE scored better than PCPs in requesting, interpreting and explaining photos during the chat, while still performing well on other clinical axes such as history-taking, diagnostic accuracy, management reasoning, communication skills, and empathy. (5/n)

Scenarios really did test multimodal clinical reasoning. While AMIE had higher diagnostic accuracy overall, accuracy dropped for both AMIE and PCPs if they did not appropriately use the images, or if the image was not of high enough quality to be interpretable (4/n)

We compared 🤖 AMIE to 👩‍⚕️ primary care physicians (PCPs) in a randomized, double-blind OSCE study with professional patient actors 🤒 using *multimodal* instant messaging 💬🖼️. Scenarios included photos of skin concerns, ECG tracings and lab tests. 🤳📈📄 (3/n)

Check out a brilliant overview of this work by co-first author @KhaledSaab11 here: https://t.co/RboE8ydvxD (2/n)

✨ New paper from our team at @GoogleAI @GoogleDeepMind - AMIE goes multimodal 👀 Our research conversational diagnostic AI now fluently interprets visual photos/tests in multimodal instant messaging. More info at: https://t.co/rg7NoefDde (1/n)

Super excited to share that our two AMIE papers by @GoogleAI @GoogleDeepMind @GoogleHealth went live on @Nature today, demonstrating a step forward for conversational medical AI!

Developing new drugs can be notoriously slow, expensive and risky. 🔬 Enter TxGemma: a more general, open and scalable set of AI models to help improve this process for researchers and developers. Find out more → https://t.co/JBXpkfhaCW

Introducing 🥁🥁TxGemma! 🧪LLM for multiple therapeutic tasks for drug development 🤏2B, 9B, and 27B 🤗Fine-tunable with transformers 🤖Agentic-Tx for agentic systems Blog: https://t.co/nTqoWgmgoF Models: https://t.co/dlcvuYdH6j