28Dec 2023
Deep Learning-Based Cardiac Arrest Risk Management System Shows Consistent Superiority in Predictive Performance

Deep Learning-Based Cardiac Arrest Risk Management System Shows Consistent Superiority in Predictive Performance

A 1-year retrospective cohort study validated the Deep Learning-based Cardiac Arrest Risk Management System (DeepCARS). The study compared DeepCARS with conventional early warning systems for predicting in-hospital cardiac arrest (IHCA) among adult patients in general wards. DeepCARS demonstrated superior predictive performance, as measured by the area under the receiver operating characteristic curve (AUROC), irrespective of Rapid Response System (RRS) operational status. The findings suggest DeepCARS as an effective IHCA screening tool for hospitals with varying RRS coverage.

  • #cardiology

Like

Save

Share