Prediction of acute hypotensive episodes using neural network multi-models

This work proposes the application of generalized regression neural network multi-models to the prediction of acute hypotensive episodes (AHE) occurring in intensive care units. Contrasting with classical auto regressive representations, multi-model schemes do not recursively use model outputs as inputs for step ahead predictions. As result, prediction errors are not propagated over the forecast horizon and long-term predictions can be accurately estimated. The effectiveness of this strategy is validated in the context of PhysioNet-Computers in Cardiology Challenge 2009. The dataset considered consists of arterial blood pressure signals, obtained from MIMIC-II Database. A correct prediction of 10 out of 10 AHE for test set A and of 37 out of 40 AHE for test set B was achieved.