Case Study: Antwerp University Hospital achieves earlier sepsis detection in vulnerable newborns with IBM AI

Healthcare startup Innocens BV uses AI from IBM to design a solution that helps detect potential signs of sepsis in vulnerable newborns

Antwerp University Hospital faced the urgent challenge of detecting late-onset sepsis in preterm newborns as early as possible in the NICU, where every minute matters and symptoms can appear hours before a diagnosis. To help address this, the hospital worked with IBM on the Innocens AI-based clinical decision support solution, designed to analyze neonatal vital signs and other patient data for early warning signs.

IBM helped implement the solution using IBM Watson Studio, IBM Cloud Pak for Data, IBM Watson Machine Learning, and IBM Client Engineering to build an explainable, federated machine learning model that runs securely within the hospital environment. The result is a 24/7 monitoring system that can identify severe sepsis cases and detect sepsis hours faster than medical staff, with clinical trials underway and plans to expand to other NICUs.

Antwerp University Hospital

David Van Laere

Founder of Innocens BV and neonatologist