Mary Meeker’s back, and her new AI report has a lot to say about healthcare. Here are 5 key slides: 1. Kaiser is all-in on AI scribes. Since late 2023, over 10,000 physicians and staff have used them to streamline documentation, already logging over 2.5M visits.

Stroke Syndromes: a radiological overview made for medical students and radiology residents (but free for everyone off course :)). #MEDed #FOAMed #RADed #FOAMrad #radres #neurorad

New paper shows a familiar result on LLMs & medicine: Doctors given clinical vignettes produce significantly more accurate diagnoses when using a custom GPT built with the (obsolete) GPT-4 than doctors with Google/Pubmed but not AI. Yet AI alone is as accurate as doctors + AI.

Another preprint 🧵: our first foray into 𝐡𝐲𝐛𝐫𝐢𝐝 𝐜𝐨𝐠𝐧𝐢𝐭𝐢𝐯𝐞 + 𝐦𝐨𝐭𝐨𝐫 𝐬𝐩𝐞𝐞𝐜𝐡 𝐛𝐫𝐚𝐢𝐧-𝐜𝐨𝐦𝐩𝐮𝐭𝐞𝐫 𝐢𝐧𝐭𝐞𝐫𝐟𝐚𝐜𝐞𝐬 Ever wished your phone knew when you perceived an error in the output it gave you (e.g., a transcription typo)? BCIs can! 1/4

We're drowning in AI assistants but still manually dragging text boxes around PowerPoint like it's 2003 😤 There's a massive gap waiting for the "Cursor of Slides"... and after testing Manus, they might just fill it. My experience 👇

Excited to share our work to be presented at CVPR this week – 𝐹𝑎𝑐𝑡𝐶ℎ𝑒𝑋𝑐𝑘𝑒𝑟, an agentic approach to fixing measurement hallucinations in multimodal generative AI models. aliceheiman Xiaoman Zhang

What if you could literally talk to your EHR? I think I would have a few choice words for Cerner…

Could this be the next generation of Thrombectomy devices?

Lingshu: A Generalist Foundation Model for Unified Multimodal Medical Understanding and Reasoning "we build a multimodal dataset enriched with extensive medical knowledge. Building on the curated data, we introduce our medical-specialized MLLM: Lingshu. Lingshu undergoes