AI Can Spot Sepsis Earlier. Why Aren't We Saving More Lives?
Published on Apr 3
AI has dramatically improved sepsis detection, achieving speed and accuracy that seemed impossible just a few years ago.
AI systems at hospitals like Johns Hopkins can now flag sepsis 4–6 hours earlier than doctors. Other systems scan hundreds of thousands of patient records for warning signs with remarkable accuracy.
But if these systems are so good at spotting sepsis early, why are 11 million people still dying from sepsis every year?
The problem isn't detection. It's what happens next.
Once doctors identify sepsis, they reach for the same toolkit they've used for decades: fluids, antibiotics, and blood pressure medications. But spotting trouble earlier only helps if we can meaningfully intervene.
This is where Galectin-3 becomes crucial.
When this protein runs wild, it triggers a cascade of inflammation and organ damage that current treatments simply don't address.
Our XGal-3® approach is fundamentally different.
Instead of just monitoring sepsis or managing symptoms, XGal-3® targets and removes the root cause—Galectin-3—from the bloodstream. This equates to stopping a fire at its source rather than just spotting it sooner.
While advancements in AI detection are a significant leap forward, the real breakthrough lies in pairing these early warnings with the treatments we are creating at Eliaz Therapeutics.
Yours in health,
The Eliaz Therapeutics Team

You're the first one here!
Early people don't usually get a prize, so here's a love letter from us.