UMC Utrecht Prediction Meetings

At the UMCU we are involved in the development and subsequent implementation of prediction algorithms. Therefore, rather than being interested in achieving the best performance on a reference dataset, we want assess the algorithms performance in an actual clinical setting. In this talk we will address two use cases for which we are currently refining the validation plan. First, we will discuss the Sleep Well Baby (SWB) project that aims to monitor sleep in preterm infants in order to schedule elective care according to sleep rhythm. Second, we discuss Big Data 4 Small Babies (BD4SB), which is intended to function as an early-warning system for the detection of late-onset sepsis (LOS) in preterm infants. The goal is for both of these systems to become available on the ward in the future.