TRENDER - A Wavelets Approach to Deriving Trends in Respiratory ICU Monitor Data

Accurate patient state assessments from the interpretation of noisy physiological time-series data are important in patient monitoring in critical care medicine. Given the large volumes of respiratory signal data originating from multiple monitoring devices in the Intensive Care Unit (ICU), we believe it is better to convert this data into trends that have abstractions increasing, decreasing or steady – this will allow medical staff to perform patient state assessments. In this paper we present the TRENDER system which is a wavelets approach to deriving trends in respiratory ICU monitor data.