Biohub's Remo-ID spots sepsis from blood cell images, 43/47 correct in early study

alex_peys · x · 2026-10-31

Biohub scientists built Remo-ID, a model running on their Remoscope imaging cytometer that recognizes sepsis from blood cell images without being told which features to look for.
Workflow: 10 microliters of blood flow into a disposable cartridge; a camera captures 2 million cells over 12 minutes while the model scans for signs of infection.
In an early study of 47 patients and controls, it correctly distinguished sepsis in 43 cases using only sample-level labels. Led by Paul Lebel; the platform was originally designed for malaria diagnosis.
Sepsis contributes to 11 million deaths yearly and lacks a single gold-standard test. The poster notes microscopy-based phenotypic screening is also proving useful for aging-related prediction tasks.

More from Apps