Tag: Patient Safety
Duke’s deterioration AI beats a vendor model across five hospitals
A two-year benchmark found Duke's deterioration index caught 55% more patient declines than a widely used vendor model at the…
Glytec aims its insulin algorithm at newborns in the NICU
The designation covers newborns and infants under two, a population with no FDA-cleared insulin dosing software.
An algorithm could triage the FDA’s 510(k) pile and catch recalls
A Management Science study found a human-plus-algorithm review model could cut device review workload and catch more unsafe devices.
One AI watchdog now covers every ward in Abu Dhabi
A year after launch, the emirate's health department says its AI patient safety platform now connects 100% of hospitals and…
FDA backs LLM juries to grade AI-written radiology reports
A $1.29M FDA contract will test whether several language models, supervised by radiologists, can evaluate AI-generated reports across roughly a…
Voice AI mispronounces one in three new drug names, benchmark finds
A new benchmark finds commercial text-to-speech models mispronounce recently approved medicines far more often than familiar ones.
Graph AI banks $13.3M to automate drug safety case work
The startup's platform cuts adverse-event case processing from hours to minutes for pharma sponsors.
12,000 voices shape UK plan for staged medical AI licenses
A doctor-led commission has handed the MHRA a roadmap that would put new medical AI models on probation before full…