Dyania Health and Nvidia are working to run Dyania’s Synapsis clinical reasoning AI inside hospital networks, using hospitals’ own servers. One deployment reviewed medical records for more than 7,800 heart-valve patients at Cleveland Clinic, drawing on up to 16 years of data to help build a model that predicts when valve replacements may fail. Clinicians remain responsible for final decisions.
Fierce Healthcare frames the effort around privacy and where data runs, saying the approach lets health systems analyze patient records at population scale without sending protected data outside their networks. The Next Web emphasizes the practical use case at Cleveland Clinic, highlighting that the AI operates on hospital infrastructure and is intended to support model development while keeping clinical oversight with doctors.
Across both reports, the focus is on bringing AI capabilities into the healthcare environment through on-premises or within-network computing, with hospitals controlling the data location and clinicians guiding interpretation and care.
