STDMn+p0: a multidimensional patient oriented data mining framework for critical care research

Abstract

In the neonatal intensive care unit (NICU) environment, critical care and treatment directly correlate to the multidimensional development of an infant and are influenced by attributes such as gender and gestational age (GA). Recent literature on guidelines developed for neonatal intensive care; do not take the gender or the GA of the infant into account. The exponential activity of a growing neonate in its early stages of life needs to be captured and embedded into algorithms designed to extract patterns of predictive temperament within the NICU domain. The STDMn+p0 framework presents an extended multidimensional approach with the ability to create patient characteristic clinical rules. Further defining NICU algorithms, through the extended use of attributes to include gender and GA, and using these new algorithms in clinical decision support systems increases the accuracy and thereby minimizes the risk of adverse events.