In 2025, patients upload scans and histories to ChatGPT, ask o3 and Grok for differentials and test suggestions, and bring the outputs to their physicians for review. Posts describe corrected diagnoses, treatment changes, and doctor-approved next steps; clinicians also trial MAI-DxO and MedGemma on open-ended cases and multimodal records within a review workflow.
🤖 ai summary based on 21 tweets
Popular demos from official product accounts, team members, or affiliated creators.
New from @GoogleResearch: The Anatomy of a Personal Health Agent. The paper explores a longstanding challenge: creating AI systems that can analyze complex health and wellness data while also providing empathetic, actionable, and evidence-based guidance. https://t.co/paNT6zIxnQ
Example of our Wayfinding AI prototype in action. It includes both the familiar multi-turn chat interface on the left, and a “best information so far” panel on the right. This two-panel interface separates the context-seeking stream from the more detailed information provision piece, enabling users to dive into the information only when they feel all relevant information has been relayed.
🚀 Big news in healthcare AI! I'm thrilled to announce the launch of OpenMed on @huggingface, releasing 380+ state-of-the-art medical NER models for free under Apache 2.0. And this is just the beginning! 🧵
Check out our state-of-the-art open weights MedGemma multimodal model for making sense of longitudinal EHR data as well as medical text and medical imaging data in various modalities (radiology, dermatology, pathology, ophthalmology, etc.) See the blog post linked below! ⬇️ https://t.co/6EqtCXl2xn https://t.co/aYlQB7loIO
We're taking a big step towards medical superintelligence. AI models have aced multiple choice medical exams – but real patients don’t come with ABC answer options. Now MAI-DxO can solve some of the world’s toughest open-ended cases with higher accuracy and lower costs. https://t.co/PVMdxIwhhx