NIH researchers presenting at NUTRITION 2026 found four popular AI photo-based calorie-tracking apps underestimated meals by 250 to 345 calories on average, using a controlled metabolic kitchen to test MyFitnessPal, LoseIt!, CalAI, and Appediet against precisely measured meals.
A Mayo Clinic AI model called REDMOD detected 73% of pancreatic cancers a median of 16 months before diagnosis by re-analyzing CT scans radiologists had called normal, according to a study published in Gut.
The CDC's Center for Forecasting and Outbreak Analytics is using AI-enhanced surveillance to assess both a Bundibugyo Ebola virus outbreak in the DRC and an Andes virus cluster aboard a cruise ship, building on research showing such systems can detect outbreaks up to 24 days earlier.
The FDA has scheduled a September 23, 2026 advisory panel review of Grail's Galleri blood test, an AI-driven screen for more than 50 cancer types, despite a UK trial earlier this year that missed its primary endpoint.
UnitedHealth CEO Stephen Hemsley faced tough questions from Rep. Robin Kelly over AI-driven claim denial rates reportedly 16 times higher than typical, as Cigna and UnitedHealth both face active lawsuits over algorithmic claims processing.
Texas A&M and USC researchers are separately developing AI models that map Alzheimer's disease progression with far more precision than current diagnostics, potentially widening the window for early treatment.
A new AI algorithm called EvORanker compares genetic evolutionary patterns across 1,000 species to identify hidden gene-disease links, helping crack a previously unsolved rare disease case.
A Boston Children's Hospital and OpenAI study used an AI reasoning model to re-analyze 376 unsolved pediatric rare disease cases, helping physicians confirm 18 new diagnoses.
UCSF researchers built a multitask deep learning model that predicts Alzheimer's diagnosis and future cognitive decline from a single baseline MRI scan and demographic data.
A Stanford Medicine pilot study found an AI tool that drafts hospital discharge summaries is safe and linked to lower physician burnout, part of a broader 2026 wave of ambient documentation AI in hospitals.