Medical signal processing in the ICU

There are currently many false alarms in the ICU. A medical monitoring system which has the ability to take into account signal data from multiple monitoring devices has the potential to decrease the number of false alarms experienced in the intensive care unit (ICU). The implementation of such a system has the potential to substantially reduce unnecessary noise and stress in the ICU, therefore resulting in both better patient care and a less stressful environment for the patient to recover in. This research explores the possibility of using an intelligent system in order to distinguish between false alarm and true alarm situations. The use of time delay phase space reconstruction of medical signal data as input to a neural network is explored. The issues regarding the phase space reconstruction and the encoding of the data for input into a neural network are discussed. Results of the phase space reconstruction and neural network implementation are presented.