More than a quarter of children admitted to the pediatric intensive care unit have or go on to develop pediatric multiple organ dysfunction syndrome (MODS). The current process for diagnosing MODS takes almost 24 hours — coupled with the fact that organ failure catastrophically impacts patient outcomes, the slow process of MODS diagnostics means the mortality rate is 10-57% in the ICU. The team will be working on developing a classifier capable of predicting MODS from a wealth of existing clinical data. ICU patients are heavily monitored, but there is no automated, scalable application that uses this captured patient’s data to accurately predict the likelihood of developing MODS in real time. We have access to existing the historical data from the physioCloud database that Drs. Winslow and Sarma have used to train a similar classifier for sepsis. The goal of the project would be to build a suitable classifier capable of detecting MODS in real time, and then building a suitable user interface to communicate results to doctors in the ICU.