earticle

영어

Intensive Care Unit (ICU) monitors generate large volumes of high frequency data from numerous cardiac and respiratory sensors attached to a patient. This presents information overload to medical staff who need to interpret this data to evaluate the physiological status of the patient at any particular point in time. In this paper we present a machine learning technique called random selection to reduce ICU data sets. We will show that this technique derives trends in ICU data sets to enable qualitative reasoning as part of a clinical decision support.