Medical AI

Zuckerberg-backed Biohub is coordinating $1.8 billion in funding and data to train AI models that could make drug research faster and more predictive.

Hippocratic AI is reported to use a 31-model safety constellation for patient-facing voice agents, raising questions about oversight and clinical deployment.

NVIDIA has released NV-Reason-CT, an open 3D CT vision-language model designed to generate radiology reports, reasoning traces, and follow-up dialogue.

NVIDIA details an agent-assisted Holoscan workflow using CLI, skills, and HoloHub to build and benchmark real-time medical AI applications.

NVIDIA has open-sourced a GPU-accelerated medical physics simulation framework for healthcare robotics, aiming to speed robot training and testing.

Midjourney announces a major expansion into medical AI, applying its generative AI technology to produce full-body ultrasound scans.

AI image startup Midjourney unveiled its first hardware product, the Midjourney Scanner, a ring-shaped ultrasonic CT device for 60-second full-body health imaging.

A new AI-powered brain-computer interface has enabled a speechless ALS patient to hold a full-time job, marking a significant medical AI milestone.

Researchers from Worcester Polytechnic Institute developed a machine-learning model that analyzes MRI scans across 95 brain regions to detect Alzheimer's disease with nearly 93% accuracy, identifying hippocampal volume loss as a key early biomarker.

Microsoft has launched Copilot Health, a dedicated AI health companion integrating wearable data from 50+ devices and records from 50,000+ U.S. hospitals via HealthEx, positioning the company's AI toward medical superintelligence for everyday patients.

Nature publishes breakthrough DeepRare AI system that uses clinical data and genetic information to diagnose rare diseases, ending years-long diagnostic odysseys for patients.

Eko Health's AI-powered digital stethoscope demonstrated 92.3% sensitivity in detecting moderate-to-severe valvular heart disease, nearly doubling the 46.2% accuracy of traditional stethoscope examinations in clinical trials.

Mount Sinai research shows AI LLMs believe medical misinformation 32-46% of the time, especially when framed as expert advice.

University of Oxford research finds AI chatbots give inconsistent medical advice, making it difficult for users to identify trustworthy health information.

Harvard researchers publish BrainIAC foundation model in Nature Neuroscience, predicting dementia, stroke, and cancer from 48,900 MRI scans.

Utah has partnered with health tech startup Doctronic to launch an AI system that will autonomously renew approximately 200 common medications after performing clinical evaluations. The initiative raises critical questions about FDA oversight authority as Doctronic claims its AI doctor doesn't require regulatory approval.

Deep learning tool provides unprecedented 3D visualization of cochlear hair cells, advancing hearing loss treatment research.

MIT researchers demonstrate that best-performing machine learning models can become worst-performing when applied to new data environments, revealing hidden risks from spurious correlations in medical AI and other critical applications.