Using AI to Predict the Onset of Sepsis

The article discusses several advancements in using AI to predict and manage sepsis, a life-threatening complication affecting millions globally. Recent developments include the TREWS system by Johns Hopkins University, which has shown promising results in early detection and treatment initiation across multiple hospitals. The University of California at San Diego has also developed the COMPOSER tool, which has reduced in-hospital sepsis mortality. Furthermore, Nanyang Technological University in Singapore introduced the SERA algorithm, enhancing early detection and reducing false positives. The article also explores the development of algorithms by Luminare, Inc., which focuses on identifying patient subtypes to better tailor sepsis interventions, aiming to reduce alert fatigue and improve clinical outcomes by providing more specific and actionable information to nurses.

Opens on Mayo Clinic Platform