Sweden's 105,000-woman MASAI trial found AI-supported mammography reduced aggressive interval cancers, while a $16 million U.S. trial called PRISM now tests the same question nationwide.
Withings' Body Scan 2, on sale in the US since July 2026 for $599.95, reads more than 90 biomarkers to flag hypertension, pre-diabetes and heart rhythm problems, part of a broader company push into reimbursed clinical care.
Garmin's new Connect+ subscription bundles an AI feature called Active Intelligence, built on decades-old Firstbeat physiological modeling, alongside CES 2026's AI food-photo tracking, testing how far a hardware-first brand can push into paid AI coaching.
Samsung's June 2026 Samsung Health overhaul adds four AI-driven metrics, Vitals, Heart Health Score, Daily Cardio Load and Fitness Index, to the Galaxy Watch, aiming to flag real problems instead of flooding wearers with data.
Abbott and Google unveiled a partnership on August 11, 2026 integrating Lingo glucose sensor data with Google's AI Health Coach, targeting 115 million Americans with prediabetes and launching one of the largest real-world metabolic health studies to date.
A radar sensor called Paul has cut resident falls by an average of 66% across 11 Arizona senior communities, but even its Swiss inventor admits he can't fully rule out that better-watched staff, not the AI itself, deserve some of the credit.
Dartmouth psychiatrist Michael Heinz built Therabot, the first generative AI mental health chatbot with published clinical trial results — but a rival psychologist warns that no algorithm can replicate the empathy real therapy requires.
A 23,132-patient study across 11 RWJBarnabas Health hospitals found an AI early warning system tied to automated rapid-response alerts cut in-hospital deaths among high-risk patients by 18%, researchers reported in NEJM AI.
Mayo Clinic's REDMOD AI model found signs of pancreatic cancer on CT scans radiologists had already called normal, detecting 73% of cases a median of 16 months before diagnosis in a nearly 2,000-scan validation study.
Boston University researchers published an antibody-specific AI framework on August 13, 2026, that improved binding-affinity predictions by up to 27% by focusing model training on the six loops responsible for antigen recognition rather than treating antibodies as generic proteins.