The role of artificial neural networks within a multi-level decision support architecture for the high dependency environment

This paper reports on a method of using artificial neural networks to impro ve the information available to clinicians, and speed of access to it, whil e monitoring patients in a high dependency environment. A multilevel modula r system architecture for a patient monitoring system is first introduced. Using recorded patient-related variables as input this structure uses defin ed processing modules to provide details on current and future patient cond ition. One method of developing these modules is to use artificial neural n etworks (ANNs), for the classifying tasks. Using cardiac surgery data, coll ected during an EU funded research project, a set of prototype ANN based mo dules are developed, evaluated and discussed.