Dyania Health's Synapsis AI Processes Decades of Patient Records to Predict Valve Deterioration Risk

·2026.11.13 22:00

Key point

The clinical reasoning engine runs on NVIDIA-accelerated computing inside hospital firewalls to keep patient data onsite while identifying care gaps.

1 / 2

Details

  • Dyania Health developed Synapsis AI, a clinical reasoning engine that connects findings across years of patient records to surface clinical signals for care teams.
  • Cleveland Clinic used the tool to analyze over 7,800 aortic valve replacement patients from 2005 to 2021, building a predictive model for valve deterioration and reintervention risk at three, five, and ten-year marks.
  • The AI completed chart review for the full cohort in a timeframe that would have taken clinicians years to complete by hand.
  • Synapsis AI runs on NVIDIA-accelerated computing within the health system's own firewall, ensuring patient data never leaves the hospital.
  • The system is built using NVIDIA NeMo open libraries and optimized with NVIDIA TensorRT and TensorRT-LLM to run efficiently on hospital GPU servers.
  • Clinicians remain at the center of decision-making, reviewing AI-generated evidence and confirming actions, while the tool handles routine chart review tasks.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.