Fried, RolandGather, UrsulaSchubert, Sigrid E.Reusch, BerndJesse, Norbert2019-11-282019-11-2820023-88579-348-2https://dl.gi.de/handle/20.500.12116/30311Physiological time series measured in intensive care exhibit trends, level changes and periods of relative constancy. This signal is overlaid with a high level of noise and many measurement artifacts, and there are dependencies between the different items measured. We develop a method which allows a reliable denoising of the data and which can separate artifacts from relevant changes in the patients condition. For clinical online application the method has to be automatized and work in real time.enMedical data analysisonline monitoringlevel shiftsoutliersText/Conference Paper1617-5468