ClinVec maps clinical codes to improve medical AI
By translating messy medical jargon into mathematical vectors, a new tool could finally fix how language models understand patient charts.
AI matches pathologists in scoring ulcerative colitis
A new foundation model matches or beats human pathologists at grading ulcerative colitis biopsies, potentially solving the clinical trial replication crisis.
Predicting breast cancer six years before diagnosis
A six-year head start on breast cancer sounds like a clinical triumph, but it comes with a massive catch.
Beyond Injection Detection: A Positive-Security Prompt Firewall that Closes the Scope and PHI Gap SOTA Classifiers Miss in Healthcare
Standard security tools fail to stop medical AI from leaking patient data because the most dangerous clinical hacks look like completely normal requests.
AI Diagnoses Brain Tumors in Twelve Minutes
A new computational pathology model exposes the limits of human sight in oncology, diagnosing complex brain tumors in minutes rather than weeks.
Simple Math Beats LLMs in ICU Shock Prediction
A new study reveals that basic physiological math outpaces complex large language models at predicting patient crash times.
AI diagnoses kidney disease with fewer labeled scans
A new active learning model proves that medical AI does not need massive, expensive datasets to outperform human-labeled benchmarks.
New benchmark exposes AI medication safety flaws
Hospital IT departments eyeing generative AI for medication safety may want to pause, as a new benchmark reveals that some top-tier models behave like hyper-anxious assistants that flag almost everything as a hazard.
Brain scans predict teen depression one year early
A new computational model uses a single brain scan to predict adolescent anxiety and depression before symptoms fully emerge.
Replacing Doctors With Code
The federal push to let autonomous AI diagnose patients and prescribe drugs is setting up a dangerous collision between political speed and medical safety.